Religion and desire for additional children in a son-preference and low fertility society: Evidence from Vietnam

Background: The association between religion and desire for additional children has 25 remained controversial and varies depending on the social context of the study. This study 26 empirically investigates the relationship between religion and desire for additional children in 27 Vietnam, a society characterized by son preference and low fertility. 28 Methods: Using nationally representative data from the 2014 Vietnam Multiple Indicator 29 Cluster Survey, the study employs Probit regression analysis to investigate the relationship 30 between religion and desire for additional children among women of reproductive age (15-49). 31 The sample consists of 5,585 women across all six economic regions. 32 Results: Religion has a strong impact on the desire to have an additional child, and the 33 relationship differs across economic regions. In three out of six regions, Christians have higher 34 fertility desire compared to non-religious people. Fertility desire of Buddhists differs from fertility 35 desire of non-religious people in two out of six regions, namely Northern Midlands and Mekong 36 River. Fertility desire of those belonging to other religions is different from that of non-religious 37 people only in Mekong River. 38 Conclusions: Identifying the impact of religion on desire for additional children can help 39 promote appropriate family planning policies. The study also demonstrates that son preference 40 exists in all regions of Vietnam, with the level increasing from the South, towards the Central and 41 peaks in the North, especially in the regions bordering China. 42


46
The world is facing a shift in fertility not only from high to low, as observed during the late 47 1950s to the late 1970s (1), but also unpredictable changes in fertility trends, especially prolonged 48 low fertility, new unions with lower demand for children, and voluntary childlessness (2). Besides, 49 due to the direct, indirect, and long-term effects brought about by changes in fertility trends, such 50 as population aging, economic consequences, and sex-imbalance due to son-preference, efforts to 51 intervene in fertility behavior are key policies for many countries (3).

52
The relationship between fertility desire and fertility is well documented (4). Fertility desire 53 is considered an important first step to achieve actual fertility (5). Thus, understanding fertility 54 desire will help policy makers to predict fertility and enable them to design appropriate policies 55 (6). There are many studies that have documented the determinants of fertility desire, including 56 economic conditions (7), health status (8), national population policy, socio-cultural context, 57 environmental impacts, and beliefs about values and benefits of children (9, 10). Religion is 58 reported to have a strong relationship with fertility desire as well as fertility in many countries (11, 59 12). For example, religious women tend to have higher fertility desire and higher fertility than 60 women without religion in the United States, Spain and most European countries (13).

61
Vietnam provides a unique case study, as the country has a special socio-political structure.

62
Under the Communist party's leadership, the majority of Vietnamese people follow the atheism 63 idea due to the policy set by the government (14). In addition, given the country's past experiences, previous studies have shown that fertility desire in Vietnam is shaped by traditions and norms such 69 as son preference (10). As previous research has shown that fertility desire in Vietnam is 70 determined by socio-cultural ideology, this study hypothesizes that religion, a belief factor similar 71 to social culture, also determines fertility desire in Vietnam.

72
In sum, religion in Vietnam is likely to influence behaviors in many ways, but goes 73 unnoticed in formulating population policies. Moreover, there has been no previous work on the 74 impact of religion on fertility behavior in Vietnam. Therefore, this study is expected to be one of 75 the first studies in Vietnam to investigate the relationship between religion and fertility desire. We 76 aim to explore how religion influences fertility desire in a son-preference and low fertility society        datasets. In this study, we merged these separate datasets, using a unique identifier, to capture 124 information both at the household level and the individual level. Given that the main purpose of 125 this study is to investigate the association between religion and desire for additional children, we 126 restricted the dataset to women with at least one child. This produced a sample of 5,585 women 127 with at least one child.

129
Fertility desire: The dependent variable is fertility desire. This is captured by women's 130 desire to have additional child in the future at the time of interview. Women who had at least one 131 child were asked "Would you like to have (a/another) child, or would you prefer not to have any 132 more children?" (0=no more, 1=have another child). We excluded women who were not physically 133 able to get pregnant.

134
Religion: Religion is the main independent variable. We employ self-reported religion,

195
The results in Table 2 suggest that son preference is prevalent across all regions of Vietnam.

196
The variable 'having at least one son' is negative and statistically significant for overall Vietnam

204
In the case of Buddhism, the results in Table 2 Table 3 show that, for Northern

209
Midlands, the level term 'Buddhism' is positive and statistically significant, while the interaction 210 term between 'Buddhism' and 'having at least one son' is negative and statistically significant.

211
The level term result implies that, for women with no son, Buddhists are more likely to want

218
For Christianity, the results in Table 2 demonstrate that, controlling for other factors,

219
Christians are more likely to want another child compared to women with no religion in the case 220 of South East, North Central, Central Highlands, and overall Vietnam. Table 3 shows that the

230
For other religions, the results in Table 2 suggest that women belonging to other religions 231 are less likely to want another child compared to those with no religion in Mekong River. For all 232 other regions and overall Vietnam, the difference is not statistically significant. Table 3 presents 233 the results including the interaction term between 'Other religions' and 'having at least one son'.

234
In the case of Central Highlands and overall Vietnam, for those with no son, women belonging to

288
Buddhists tend to want more children compared to those with no religion. On the other hand, in

289
Mekong River Buddhists are less likely to want another child compared to those with no religion.

290
The combination of the differences in the sectors of employment, the level of son preference and

310
Another interesting finding is that the Buddhists have stronger son preference than those 311 without religion in the Northern Midlands. In Vietnam, non-religious people are likely the group 312 that work in government agencies because non-religiousism is the direction of the communist 313 regime (31). They also tend to have higher education which is required for government positions,

314
and high educational attainment has been linked to low fertility in many studies (32, 33).

315
Additionally, government workers have pension funds, so they do not need to financially depend 316 on children during old age. High education and old age financial security can lead to low fertility 317 desire among government workers. The one-or-two child policy also applies only to those Roman Catholic marriage and against the Protestants' bible content "be fruitful and multiply" (35).

328
Besides, for those with no son, Christians are more likely to want another child compared to 329 women with no religion. As discussed above, non-religious people tend to work as government 330 officials, therefore the one-or-two-child policy can be an explanation here. On the other hand,

331
Christians are free to have more children until they have a son.

332
In addition, there are two conflicting trends in the fertility desire between Christians and higher education and income can lead to higher fertility due to the ability to bring up children (36).

353
It is also possible that the beliefs of women belonging to other religions influence their fertility

Conclusions
Our research provides key findings on the relationship between religion and desire for 366 additional children in Vietnam, a society characterized by son preference and low fertility.

367
Religion has strong impact on desire to have an additional child, especially for women with low