North–South discrepancy and gender role attitudes: evidence from Vietnam

In Vietnam, it is commonly believed that gender norms, sex labour segregation, and structural organization of social institutions often favour male dominance while restricting women’s roles in domestic spheres. However, there is a scant literature of Vietnamese scholarship on the determinants of gender role attitudes, especially geographical disparities. This paper aims to fill this void by using a nationally representative survey with 8288 respondents. Our findings suggested that age, marital status, religion, education, living area, region, ethnicity, and personal monthly income are the factors that predicted gender attitudes. In terms of regional disparities, we found that Northerners were more permissive in their gender attitudes than Southerners, which may be explained by distinct historical and political trajectories in Northern and Southern Vietnam during the last century. There were, however, inconsistent patterns among different age cohorts whereby region significantly impacted the attitudes of women born before the end of the French War in 1954, men born after the Reunification in 1975, as well as both men and women born between 1954 and 1975.


Introduction
The concept of gender role attitudes can be defined as the views towards the roles of men and women in varying areas, including the labour division in family life, the workforce, or politics (Halimi et al. 2021;Philip et al. 2019;Oskamp and Schultz 2005). There is a substantial body of scholarship exploring gender role attitudes in Western countries since the second half of the twentieth century, which overall indicates a substantial change in gender relations (see, for instance, Loo and Thorpe 1998;Neve 1995;Spence and Hahn 1997). This transformation is believed to have resulted from the rise of postmodernism and individualism, gender equality movements, increased women's participation in the workforce, changing demographics of family, as well as the introduction of more open laws and norms (Kien 2018;Bolzendahl and Myers 2004).
Among many factors affecting gender attitudes, regional difference has been considered as a significant determinant as it affects attitudinal support towards gender roles across subgroups, such as ethnicity and race (Powers et al. 2003;Suitor and Carter 1999;Suitor et al. 2001;Twenge 1997). It is, however, less clear as to how impact of the region factor differs by age cohort. Additionally, while most literature addressing region's effect on gender role attitudes has been examined in Western settings, it is relatively unknown whether similar findings might apply to non-Western societies. In the South East Asian country of Vietnam, gender roles were dominated by traditional gender norms, sex-labour distribution, family tradition, which clearly separate between women's domestic roles and men's duties and obligations in the family, the workforce, and the public spheres (see for instance, Bélanger and Khuat 1996). It is commonly believed that there is a regional difference in gender role attitudes between people of Northern and Southern Vietnam. This belief is based on the perception that Northern and Southern Vietnam had been exposed to different sociopolitical conditions throughout the Vietnam War . This paper aims to test this common knowledge by investigating the attitudes of Vietnamese men and women in the North and in the South towards gender roles in multiple domains across three age cohorts: (1) born before 1954; (2) born between 1954 and 1975; and (3) born after 1975. Within the scope of this paper, we accept the limitation that this classification could contain some bias factors since there are many different social groups being born after 1975. In the sections that follow, we review the theoretical literature to provide an understanding of how location variables can impact gender role attitudes. This is followed by the section on the gender status in Vietnam. We then describe in detail the survey and dataset upon which our study was built, as well as variables used in our predicting models. Subsequently, we present the results of our analysis, including a mean comparison and the findings regarding the effect of region variable on gender role attitudes across different age cohorts. Finally, we discuss our results in reference to the existing body of literature before providing suggestions for future research.

Location as a predictor of gender role attitudes
Previous research portray several consistent patterns regarding the significance of region in determining attitudes towards gender roles. For instance, studies by Hurlbert (1989), Twenge (1997), Bolzendahl and Myers (2004), and Carter and Borch (2005) using a longitudinal data set across the United States reveal the distinctiveness of the South when it comes to feminist views. According to them, Southern Americans tend to hold more conservative or traditional attitudes towards feminism as compared to non-Southerners. The regional divide, as maintained by existing scholars, is attributable to variations, such as in religion (Ellison and Musick 1993), education, rural residence, and political favouritism (Griffin and Hargis 2008).
In this study, we anticipate the different effects of region in predicting gender role attitudes given the distinct historical and political trajectories that exposed Northern and Southern Vietnam to different degrees of socialist experience in the last century.
Existing studies have pointed out that social norms and viewpoints, including gender egalitarian attitudes are differently shaped by communism/socialism and capitalism (Alesina and Fuchs-Schündeln 2007;Bauernschuster and Rainer 2012;Breen and Cooke 2005;Beblo and Goerges 2015;Lippmann and Senik 2018;Lippmann et al. 2020). For instance, using longitudinal data from Germany's General Social Survey, Bauernschuster and Rainer (2012) argue that the socialist regime established in East Germany made Easterners more likely to show their egalitarian gender role attitudes towards work compatibility, motherhood, and the appropriateness of men and women roles as compared to West Germans. This gap in gender role attitudes has not disappeared but persisted and even significantly widened after the reunification. The lasting effect of socialism on gender role attitudes is similarly found in a more recent study by Lippmann et al. (2020) whose analysis was based on the German Socio-Economic Panel from 1991 to 2012. Their findings show that in the 41-year division of the country, East Germany reigned by socialist parties adopted gender-equalising policies which encouraged women to participate in the labour market. Meanwhile, under the capitalist regime West Germany maintained their traditional family policy which rested on the male-breadwinner norm. At the time of reunification in 1989, the study reports a notably higher rate of women participating in the work force in East Germany (89%) than in West Germany (56%). The study further reveals that a woman in West Germany is more likely to withdraw from the labour market if her potential income is higher than that of her husband, while an East Germany woman is likely to remain in the labour force because women value paid work as much as men do. Since there is limited discussion on how communism and capitalism could influence gender role attitudes, the above-mentioned example of Germany provides a good referencing point for our study.

Gender differences
Early scholarly discourses by Chodorow (1999) and Gilligan (1982) and books by Tannen (1990) and Gray (1992) have suggested that attitudes of men and women 1 3 differ, especially towards social phenomena, which is consistent across different country settings (Frieze et al. 2003;Olson et al. 2007). It has been argued that men and women, as well as boys and girls, will interpret the characteristics and meanings of social situations differently despite being exposed to the same information (Suitor et al. 2001). Much of the existing empirical research has also been devoted to documenting the attitudinal differences regarding gender roles between women and men. A common finding in these studies is that women appear more liberal in attitudes (Bolzendahl and Myers 2004;Hunter and Sellers 1998;Gibbons et al. 1991;King and King 1990;McHugh and Frieze 1997;Mason and Lu 1988;Neve 1995;Wilkie 1993) and that men are more resistant to gender equality within the family and the workplace (Mason and Lu 1988;Twenge 1997). Attitudes of women are also found to be more egalitarian in areas related to broader societal contexts, for instance in political arenas (Rice and Coates 1995). It is worth noting that sex differences are not always gender differences because the latter is a result of social construction and relations between men and women who occupy distinct roles and carry different responsibilities in society. As such, while the literature discussed above might be useful as a starting point, we also incorporated more relational roles to identify the gender differences in the context of Vietnam. For example, we based our gender role attitude variable on 17 questions covering various aspects of gender roles and values such as sexuality, marriage and son preferences, family decision-making, education, and leadership.

Age cohort differences
Glenn (1977) defines a cohort as "those people within a geographically or otherwise delineated population who experienced the same significant life event within a given period of time" (p. 8). According to Ryder (1985), the coherence and continuity of each birth cohort results from its distinctive constituents as well as its 'macroanalytic' features. Changes in formal education, peer group socialisation, and personal historical experience characterise successive cohorts.
There has been ample evidence illustrating the cohort effects related to the social and historical environment on how women and men are socialised differently about gender roles (Brewster and Padavic 2000;Davis and Robinson 1991). For instance, when studying the shift in gender ideology in the United States from 1977 to 1996, Brewster and Padavic (2000) suggest significant cohort differences in the determinants of gender role attitudes. Three variables in their study, including sex, education, and church attendance, differed significantly across birth cohorts. Specifically, sex differences in attitudes were wider among members of the youngest cohorts (those born between 1945 and 1978). In terms of education, their analyses reveal that the relationship between education and gender role attitudes was more pronounced for cohorts born before 1945 relative to younger generations. Meanwhile, the church attendance effect appeared markedly larger for the more recent cohort. Nevertheless, previous empirical research has not sufficiently revealed as to how age cohort differences in gender role attitudes might vary for those living in different regions.

Background on gender status in Vietnam
There is a claim that Confucianism, which was imported in Vietnam since the tenth century AD during the Chinese occupation (Zhang and Locke 2001) and fostered conservative gender relationships, has heavily influenced Vietnamese society (Grosse 2015). Others, however, argue that gender inequality in Vietnam is more the result of the country's political systems and social structures of rural villages and families, which created and reinforced a patriarchal system that portrayed women as a subservient group to men (for instance, Tetreault 1996).
The French colonisation since 1858 with their introduction of Western gender values challenged the traditional gender ideology (Tetreault 1996). Yet, it was not until early 1930s with the establishment of the Indochinese (later Vietnamese) Communist Party and Women's Union when the concepts of gender and gender status started to change in Vietnam (Que 1996). The Communist Party of Vietnam (CPV) assumed power and became a ruling party in the North of Vietnam in 1945 where Communist ideals were initially espoused (Tucker 1998). Following the declaration on the formation of New Democratic and Republic State of Vietnam, the first Constitution was introduced in 1946 which officially states 'equality between men and women' (Fahey 1998: 225). Later on, the CPV repeatedly highlighted the goal of gender equality after the 1975 Reunification (Knodel et al. 2004) and maintained the principles of 1946 Constitution in the 1980 Constitution-the first Constitution of the reunified Socialist Republic of Vietnam through its gender and political agendas (Eisen 1984).

One country, two regimes
This section aims to provide a very general contextualisation of Vietnam's historical division into two regimes, which helps to make sense of the research model and findings in this study. The twentieth century witnessed momentous upheavals in Vietnam, including the eradication of feudalism, several decade-long wars, economic centralisation, and trade liberalisation, which led to radical transformations across the whole country. Nonetheless, war experiences and recent shift to a marketdriven economy (since 1986) differentially affected upon the political, economic and social histories of Northern and Southern regions (Keyes 1995).
From the end of the French war in 1954 until the 1975 reunification, the 17th Parallel was determined as a boundary that divided the country into two different regimes. In the north of the Parallel, the people began the country's reconstruction and building a socialist state under the communist government of the Democratic Republic of Vietnam. Meanwhile, U.S. neo-colonialism was started below the 17th Parallel as a new form of foreign domination (Eisen 1984). In the Communist North, traditional gender roles were challenged as millions of Northern Vietnamese women actively participated in the war in the local guerrilla and militia forces, as well as in other professional capacities (Anderson 2010). Their new roles as the long-haired warriors redefined the standards of womanhood from a domesticated life to their historically unprecedented responsibilities on the battlefield to fight for their country (Taylor 2007). Meanwhile, the consciousness of the Southerners had been exposed to Western gender norms under neo-colonial control, whereby some authors believed that the people were instilled with the notion of democracy that meant "the freedom to wear miniskirts and ride Honda motorcycles," which differed from the socialist society in the North (Eisen 1984: 83-84).
There were improvements following the 1975 Reunification when the socialist regime was imposed on the South; however, the level of consciousness remained unequal between the two regions. As pointed out by Eisen (1984) and Werner (2009), women in the North had higher literacy, better education, childcare, and other social facilities, as well as higher participation in the labour force and political organisations. The collectivisation of household businesses and the encouragement of women's participation and representation there contributed to changes in gender role attitudes (Eisen 1984). Data from the 1976 National Assembly election revealed the North-South discrepancy in women's leadership aspects, showing that only 23% of the Southern nominees were women in comparison with 32% of Northern candidates (Eisen 1984: 46). In the South where socialist gender equalisation programs only started in 1975, efforts of the CPV to spread their ideas were often met with lower degree of compliance (Goodkind 1996).

Gender situation since Doi Moi
Rapid economic and social transformations since the 1986 Doi Moi (Renovation) have contributed to expanding educational and working opportunities for women, enhancing their well-being, and advancing women's status domestically and in public spheres. The government of Vietnam continues to show its commitment to promoting gender equality as evident in a number of national laws, the ratification of key international conventions, as well as the incorporation of gender goals in national socio-economic development strategies and plans (UN Women 2016). Significant progress in narrowing gender disparities in many dimensions has been reported. For instance, gender gaps have been closed at all educational levels (UNDP 2015). With 24% of the total National Assembly seats occupied by women, Vietnam records one of the highest rates of women's political participation in Asia. Women's labour force participation is also high, at 73% (GSO 2015). A decline in traditions upholding gender disparities, such as arranged marriage, child marriage, polygamy, and patrilineality has been observed; and simultaneously, an increase in political participation among women (Goodkind 1995;Volkmann 2005;Rydström et al. 2008;Werner 2009; World Bank 2011).
Regional variations, however, have persisted between the North and the South in different aspects, for instance, regarding marriage rituals (Goodkind 1996), household living arrangement (Truong et al. 1997), household composition (Bélanger 2000), and premarital sexual behaviour (Ghuman et al. 2006). Regarding labour force participation, according to the Labour Force Survey 2020 (General Statistics Office (GSO) 2020), the national gender gap in labour force participation rate is 10.9 percent, but varies notably between regions. Among six socio-economic regions, the gap is the least pronounced in the Northern midlands and mountain areas (4.1 percent), followed by Red River Delta (6.6 percent). Meanwhile, the widest gap is recorded at 19.4 percent in the Mekong Delta in the South, followed by the South East (15.1 per cent) (GSO 2015). There remain pervasive gaps between men and women in various areas such as access to decent jobs, public infrastructure and services (UN Women and GSO 2020), access to public space, freedom of speech, the occupation of assembly's seats, and leadership positions (OECD 2014), as well as division of household chores (Teerawichitchainan et al. 2009).
Despite the prolific literature on regional stratification since Doi Moi, our knowledge is limited regarding gender role attitudes between the two regions. The political, social, and historical differences between the North and South of Vietnam lead us to believe that there are regional differences in attitudes towards gender roles. Given the North's greater exposure to Communist ideologies over a longer period (since 1954) than the South (since 1975), we hypothesise that people living in the North would appear more permissive in attitudes towards gender roles in comparison to their counterparts in the South. In addition, after reviewing radical changes in various aspects of society in the post-war era, notably the increased social status of Vietnamese women, we also expect to see some pattern of attitudinal differences by age cohort, as well as by sex. Our study therefore seeks answers to the three following questions: (1) What constitutes Vietnamese people's gender role attitudes?
(2) What are the differences between men and women's perceptions of gender roles? (3) Are there any differences in attitudes between age cohorts in the Northern and Southern regions?

Data collection
This study used data from the national survey Social Determinants of Gender Inequality in Vietnam completed in 2015 in nine cities and provinces, including Hanoi, Thai Binh, Phu Tho, Da Nang, Lam Dong, Binh Thuan, Ho Chi Minh City, Vinh Long, and Tay Ninh (see Do et al. 2018). In each province, lists of wards/communes and households were obtained from the Provincial Statistics Office. A multistage probability sampling method was applied to ensure the sample's representation of the whole population.

Dependent variable
We constructed the dependent variable gender role attitudes by summing the responses to 17 questions specifically designed to tap into individuals' stances towards gender egalitarianism (Appendix 1). Respondents were asked to locate themselves on a three-item scale, ranging from agree (coded 1), partially agree (coded 2) or disagree (coded 3) with certain statements on the role of men and women in five domains: (1) sexuality; (2) marriage and son preferences; (3) family decision-making; (4) education and (5) leadership. Possible scores range from 17 for those holding anti-egalitarian attitudes on all questions to 51 for respondents who consistently endorsed egalitarianism. The Cronbach's alpha of the scale equals 0.78, demonstrating a high internal consistency.

Independent variable
Region variable. We grouped respondents who lived in three provinces of Hanoi, Thai Binh, and Phu Tho into the North region (coded 1) while classified the remaining provinces as 'South' (coded 0). This division also overlaps with the previous segregation of Vietnam into two zones from 1954 to 1975 when the North and South of Vietnam were determined upper and lower the 17th Parallel. While there are three provinces in the North and six in the South, this difference can be mediated by the large sample of the two cohorts.
Socio-demographic variables. Guided by theoretical perspectives reviewed in earlier literature (Bolzendahl and Myers 2004;Davis and Greenstein 2009;Powers et al. 2003), we included a variety of social and demographic control variables in our multivariate analyses. These variables include age, educational level, marital status, employment status, living area, family size, personal monthly income, religion and ethnicity. Age of respondent was calculated until 2015-the year of the survey. Respondents were asked about the highest educational degree they had obtained, whose answers were then coded as Primary School or lower (coded 1); Lower Secondary Diploma (coded 2); Upper Secondary Diploma (coded 3); Vocational College Degree (coded 4); and University/Higher Degree (coded 5). Marital status at the time of the survey was inquired and coded as Married = 1; Not Married (never married, separated/divorced, or widowed) = 0. Respondents were also asked: whether they were employed (coded 1) or not (coded 0); their ethnicity (Kinh = 1, Minority = 0); and religion: Atheist (coded 0), Buddhist (coded 1), Others (coded 2). Their personal monthly income was also inquired and then grouped into three ranges: Below 2,000,000 VNDs (coded 0); From 2,000,000 to 5,000,000 VNDs (coded 1); and Above 5,000,000 VNDs (coded 2). 2 Whether respondent lived in urban (coded 1) or rural area (coded 0), and the number of members living in their household were also controlled in our models.
We conducted analyses over the whole sample and segregated between women and men across three age groups. In total, our sample included 8,288 respondents (4,128 men and 4,160 women) with an average age of 42.56 years (SD = 12.10, range = 18-65). When divided by age cohort, it was found that 11.97% (n = 992) of the respondents were born before 1954; more than half (n = 4,636; 55.94%) were born between 1954 and 1975 while the rest (n = 2,660; 32.09%) were born after 1975-the end of the Vietnam War. People living in the North accounted for nearly a quarter of the whole sample (n = 1,854; 22.37%), with the remainder residing to the south of the former 17th Parallel (n = 6,434; 77.63%). Bivariate correlations of all the variables we analysed in multivariate regression models are shown in Table 1.

Mean scores of gender role attitudes
We began by looking into gender attitude scores separately by gender, region and age cohort. The overall mean score of gender role attitudes for the whole sample was 31.71 (SD = 5.65) while the respective mean scores for women and men were 31.34 (SD = 5.58) and 32.08 (SD = 5.71). Comparing scores for each subgroup are presented in Table 2. Generally, people living in the North showed higher scores than their Southern counterparts (statistically significant difference in means at p < 0.01), which suggests that Northerners' attitudes towards gender roles were more equal. By age cohort, as expected, respondents of the youngest subgroup appeared to hold the most egalitarian gender role attitudes. Such patterns remained when we disaggregated by gender. Among the women subsample, Northern women born after 1975 expressed the most egalitarian attitudes; meanwhile, the lowest score was recorded for those born before 1954 residing in Southern region of Vietnam. Similarly, within men category, Southerners born before 1954 seemed to be the most traditional in gender role attitudes while Northern men born after 1975 were the most egalitarian. The average score of the youngest subgroup comprising men in the North was also the highest among the whole sample.
Contrasting many previous studies, our findings demonstrated that men were more egalitarian in gender role attitudes as compared to women (statistically significant difference in means at p < 0.01). This finding holds true not only in the total surveyed population, but also when intra-cohort or intra-region comparisons were considered.

Regional effects on gender role attitudes across different age cohorts
We then continued to explore correlates of gender role attitudes and specifically the effect of region as a predictor across the whole sample (Model 1), as well as among the women (Model 2) and men (Model 3) categories. Considering the whole surveyed sample as shown in Model 1 (Table 3), overall, age of respondent, educational level, marital status, ethnicity, religion, living in urban area, monthly income, and region significantly determined gender role attitudes. Specifically, higher educational attainment levels predicted greater scores on gender role attitudes. Concerning religion, respondents who were practitioners of a religion other than Buddhism were more equal in their attitudinal support. In terms of living area, urban residents experienced more egalitarian attitude scores than those in rural areas. As regards personal income, respondents who could earn more than 5 million VNDs a month showed higher scores than earners who made less than 2 million. Meanwhile, egalitarian attitudes decreased with an increase in age, meaning that older respondents exhibiting more traditional stances towards gender roles than younger people. Being married at the time of the survey or originating from Kinh group (the majority ethnicity) diminished the level of gender egalitarianism. Our findings did not support Table 1 Correlation matrix of all variables Neither variables of being employed nor average number of household members proved significant in our models. The significance of independent variables, nevertheless, varied between women and men subgroups. Regarding attitudes of women (Model 2), gender role attitudes were predicted by marital status, which was not observed among men: given that all other variables remaining constant, married women reported less egalitarian gender role attitudes as compared to those who got divorced, separated, or never got married. Regarding income, earning a monthly income of 2 to 5 million VNDs and over 5 million VNDs were both significant predictors of gender attitudinal support among women (p < 0.05 and p < 0.01 respectively). Conversely, income did not show any significance in the model predicting men's attitudes (Model 3). In this model, both religion-related variables were found as significant predictors, although in opposite manners. Accordingly, as compared to those who did not belong to any religion, Buddhist respondents reported less equal attitudes. Meanwhile men claiming themselves as followers of other religions were more permissive in their attitudinal support than men who were Buddhist or Atheist.
The region effect had a significant part in explaining gender role attitudes in the whole sample, as well as within categories of men and women. Accordingly, as compared to people residing in the South, those living in the North appeared more egalitarian. The magnitude of the region's coefficient was marginally larger when predicting the gender attitudes of women than those among men.
We conducted further sex-disaggregated analyses to examine various determinants of gender role attitudes within each age cohort (Model 4 to Model 6 predicting the women's gender attitudes and Model 7 to Model 9 predicting men's gender attitudes). Findings presented in Table 4 illustrate that the region variable became an important predictor among the most senior women cohort who were born before were not predicted by region as can be seen in Model 6. Analyses on men subgroups reveal a different pattern. Dissimilar to the eldest women age cohort, region did not show any significant impact on gender role attitudes of men who were born before 1954 (Model 7, p > 0.05). In contrast, the attitudes towards gender egalitarianism of the two younger age cohorts were Adjusted R-squared 0.16 0.18 0.14 significantly explained by the region variable (Table 5). Model 8 predicting the attitudes of men born between 1954 and 1975 and Model 9 exploring the attitudes of those born after 1975 both displayed that residing in the North increased the gender egalitarian level (p < 0.05 and p < 0.001, respectively). The region effect was noticeably larger in the youngest age cohort born after 1975 than among those who were born earlier in the period from 1954 to 1975.

Discussion and conclusion
This paper has shed light on the factors predicting gender role attitudes and particularly on the variations in attitudinal support towards gender roles by living region, sex, and age cohort among a nationally representative sample of Vietnamese respondents. Overall, we found that respondents with the most egalitarian our results demonstrated that people of younger ages, with better educational levels, living in urban areas, following a religion other than Buddhism, or earning higher incomes reported more permissive attitudes towards gender roles. Meanwhile, married respondents or those of the ethnic majority were those with more traditional attitudes.
The present study pointed out that Vietnamese women were less progressive in their attitudes towards gender roles as compared to their men counterparts, which was unusual not only in earlier scholarship from the West (Bolzendahl and Myers 2004;Hunter and Sellers 1998;King and King 1990;Mason andLu 1988;McHugh and Frieze 1997;Neve 1995), but also in other Asian societies (Qian and Sayer 2016;Tu and Pei-Shan 2005;Uji et al. 2006;Zuo et al. 2012). Our finding, nevertheless, corresponds with a previous research by Knodel et al. (2004) studying 1,296 married persons in the Red River Delta of Vietnam, which also revealed that men were more likely to support gender equality within household than women. Knodel et al. (2004) argue that this reflects the awareness of the social ideals advocated by the state and other organisations, while such awareness does not necessarily turn into practice. Vietnamese men probably want to inhibit their inequality by attempting to conceal their true attitudes. Meanwhile, women, as remaining confined by patriarchal hierarchy, might still suffer from their own prejudices and lack the confidence to assign themselves an equally important status with men. Despite substantial progress on women's advancement, various scholars claim that the patriarchal thinking or other Confucianism-influenced ideologies still have a central role in shaping public discussion surrounding the modern image of womanhood in contemporary Vietnam (Binh 2004;Drummond and Rydstrøm 2004;Rydström et al. 2008). Women's more traditional attitudes as our results suggested might also be attributed to state movements and campaigns of post-Doi Moi, especially through the Women's Union, which attempted to re-essentialize femininity surrounding traditional maternal and domestic roles (Khuat et al. 2010;Hoang 2020). As captured elsewhere, women's increasing engagement in the labour market and political arenas provoked anxieties over the long-held gendered order, leading to systemic interventions to bring back traditional values in the post-socialist gender regime (see Leshkowich 2011).
Our analysis highlighted that region was a significant predictor of gender role attitudes; however, the effects of this variable were inconsistent among different age cohorts. As for women, region significantly explained gender role attitudes of two age cohorts, including those who were born before 1954 and from 1954 to 1975. However, the region effect disappeared when analysis on the youngest age cohort was considered. Including control variables such as education probably weakened the region effect among these population subgroups since educational attainment has become more homogenous across regions for younger age cohorts. Compared to seniors born before the Vietnam War, women born after 1975 received far better education and information exposure. They are also more likely to participate in the labour market and can move more easily across regions. The rising educational attainments, greater mobility, and information exposure, as 1 3 well as expanding job opportunities for women all over the country might contribute to lessening the impact of region factor in our predicting models.
As for men, we found no significant effect of region in the most senior subgroup comprising those who were born before 1954. For this eldest age cohort, demographic characteristics and influence of feudal ideologies were rather similar across different regions, which might explain the insignificance of region as a predictor of gender role attitudes. Regarding the two younger age cohorts, our study found out that men living in the North reported higher levels of egalitarianism than those residing in Southern regions.
These findings support our initial anticipation that Northern men's longer exposure to socialist ideals along with the associated gender messages and policies could explain greater awareness of feminism while challenging conservative Confucian teachings that suppress women's status. These significant regional differences contradict what Jayakody and Phuong (2013) reveal in their study which analysed data of the Vietnam Family Survey. In their research, individuals in the South were more likely than Northerners to disagree with the statement that men should make important decisions in the family. Further, their study shows that Northern people have lower probability of endorsing the equal roles of husbands and wives when it comes to sharing housework in comparison with those in the South. However, our result shares with the findings by Teerawichitchainan et al. (2010) who found greater involvement of Northern husbands in most domestic tasks than Southern counterparts when using two surveys with a sample of 2592 respondent. This study also reports the cohort differences whereby men belonging to the most recent marriage cohort are more likely to contribute to budget management within their household as compared to those in the two earlier cohorts.
This current study has several limitations. First, we used living region variable to determine the respondent's origin which could not fully reflect their immigrant status. Additionally, some factors such as the spouse's income or employment status which have been found as important predictors of gender role attitudes were absent in our explanatory models due to data unavailability. Future work should aim to collect further data to better understand how those variables might affect the gender role attitudes among the public. A longitudinal study is also recommended to investigate whether the current regional differences remain the same over time. Despite its shortcomings, our results are consistent with earlier research works in exploring the determinants of gender role attitudes among men and women. It contributes to addressing a literature gap regarding the relationships between region, gender, and age cohort. Pointing out the differences in region effect across subgroups, we emphasise that the impact of region might be obfuscated without including gender and age cohort in predicting models. Notes 1. The past Parallel is currently lying within the province of Quang Tri. 2. Exchange rate at the time of analysis: 23,200 VND is approximately 1 USD.