Parents’ Valuation of Approving a Child’s Spouse in a Context of Marital Change

Abstract

Previous research on arranged marriage focuses on young people getting married. In this article, the author turns the focus toward parents, examining whether Nepali parents value approving their child’s future spouse. A fifth of parents did not value approving their child’s spouse. They then investigated which parents value such approval less than others and how these connections differ by gender. Parents who had low-quality relationships with their children and believed it is better, in general, for children to choose their own spouses were less likely to value approval. Parents of children with salaried work experience, especially mothers of daughters, were also less likely to value approval. By contrast, children’s education, parents’ marital experience, and valuation of parental respect were not important. Believing arranged marriage and respect for parents will decrease in the future and that parents relinquish spouse approval in exchange for old age support were also unimportant.

Keywords

marriage intergenerational ideational developmental idealism Nepal

Historically, arranged marriages, in which parents selected spouses for their children, were common in many parts of Asia and Africa. In recent decades, however, young people have become increasingly active in choosing their own spouses. Surveys from several contexts document declines in arranged marriage (Allendorf & Pandian, 2016; Malhotra, 1991; Meekers, 1995; Nedoluzhko & Agadjanian, 2015). For example, in Nepal, the proportion of individuals who participated in the selection of their spouse rose from nearly zero among those married in 1936-1945 (Axinn, Ghimire, & Barber, 2008) to three quarters of those married in 2008-2012 (Allendorf, 2017). Ethnographic studies also point to marital change in many other localities (Harkness & Khaled, 2014; Hirsch & Wardlow, 2006; Marsden, 2007; Rebhun, 1999).

Much of the research on arranged marriage, or spouse choice, focuses on the young people at risk of marriage. It is the characteristics of these youth that are highlighted as sources of change and variation in marriage. Convergence theory, for example, identifies young people’s independent incomes in industrial, urban societies as a key driver of greater freedom of mate choice (Burgess & Locke, 1960; Goode, 1963). More recently, Ghimire, Axinn, Yabiku, and Thornton (2006) highlight young people’s broader nonfamily experiences, including not only work, but also media exposure and participation in youth clubs. And, Allendorf and Thornton (2015) point to young people’s endorsement of developmental idealism, a collection of values and beliefs that identify some family behaviors as inherently good, modern, and related to socioeconomic development. Many ethnographies also identify young people’s desire to marry someone they love, or someone they themselves choose to marry, as a key driver of self-choice, or “love,” marriages (Ahearn, 2001; Kaur & Palriwala, 2014).

The role of the potential arrangers, the parents of prospective brides and grooms, have received less attention. In some studies, parents hardly appear, while in others they are portrayed as a source of considerable opposition to children’s attempts to marry their chosen partners (Chowdhry, 2007; Cole & Thomas, 2009; Mody, 2008; Netting, 2009). In some cases, these struggles involve police and judicial courts and even end in suicide and honor killings (Mody, 2008; Rocca, Rathod, Falle, Pande, & Krishnan, 2009). However, parents are exposed to many of the same social, economic, and ideational changes and pressures that their children face. While some parents act as barriers to greater choice among their children, others may ambivalently or happily relinquish control. Many young people may also be reluctant to choose their own spouses without the support, or at least the acquiescence, of their parents. In turn, parents’ valuation of approving a child’s spouse is not only important in its own right, but may be an important source of change and variation in marriage behavior.

In this article, I build on the previous literature by turning the focus squarely toward parents and examining their valuation of approving a child’s future spouse in Nepal, a context experiencing a decline in arranged marriage. Specifically, I investigate variation in parents’ valuation of spouse approval and identify whether a sizeable proportion of parents do not value it. Next, I examine which parents value approval less than others. I evaluate whether parents’ endorsement of developmental idealism about arranged marriage and parental respect, the quality of the relationship with their child, their own marital experiences, belief in strategic tradeoffs, and children’s (potential) independence are important correlates of parents’ valuation of spouse approval. These connections are likely to differ between mothers and fathers and between parents considering the marriage of daughter versus a son. Thus, I also examine whether these patterns and connections differ by gender of parents and their children.

Background, Conceptualization, and Hypotheses

Theories of variation and change in marriage, as well as family more broadly, point to a variety of socioeconomic and ideational influences on marriage. Studies on South Asia reinforce many of the factors identified in sociological theory, while also pointing to factors that are particularly relevant to the South Asian context. Drawing on both types of sources, I identified several factors that are likely connected to parents’ valuation of spouse approval. Below, I describe how each of these factors may shape parents’ valuation of spouse approval. It is important to note that this is not a comprehensive list; it is likely that there are other factors related to parents’ valuation of approving a child’s spouse.

Developmental Idealism About Arranged Marriage and Parental Respect

Developmental idealism is a collection of values and beliefs about socioeconomic development and family (Thornton, 2001, 2005). The content of these values and beliefs can be summarized in two main propositions. First, societies occupy different levels of development and generally move from “traditional” to “modern.” And, second, “modern” family behaviors are not only inherently good but causes and consequences of development. The origins of developmental idealism lies with Northwest European elites who anthrocentrically viewed their own societies as the most developed and modern. Thus, characteristics of Northwest European societies and its diaspora, like low fertility and nuclear families, were identified as good, modern, and causally related to development. Conversely, characteristics of societies found largely outside Northwest Europe, like high fertility and extended families, were identified as traditional and unhelpful to development.

Developmental idealism theory posits that the spread and adoption of this collection of values and beliefs leads to changes in family behavior (Thornton, 2001, 2005). Specifically, the theory suggests that developmental idealism spread from Northwest European elites to ordinary people around the world via colonialism, schooling, mass media, migration, foreign aid, government policy, and other mechanisms. As people encounter and adopt developmental idealism, they adopt “modern” behaviors. They are motivated to do so because they view “modern” behaviors as not only inherently desirable but a means for securing a better life and development more generally.

Within developmental idealism, arranged marriage and high levels of parental respect are “traditional” behaviors expected to decline in favor of the “modern” behaviors of self-choice marriage and youthful autonomy (Allendorf & Thornton, 2015; Thornton, 2005). This identification is due to the customary geographic distribution of these behaviors. In Northwest Europe, youth customarily had high levels of autonomy from parents and chose their own spouses, often on the basis of affection developed through a period of courtship. By contrast, arranged marriages, in which parents chose spouses for their children, and high levels of parental respect and control more generally, were customary in many Asian and African societies.

Parents’ valuation of approving a child’s spouse is itself a developmental idealist schema. Specifically, viewing spouse approval as inherently undesirable or unimportant is consistent with developmental idealism. Since such developmental idealist schemas are part of a collection, such a view is likely connected to other values and beliefs found within developmental idealism. For example, parents who believe that arranged marriage will decline over time as a society becomes richer likely do not value approving their child’s spouse.

It is also important to distinguish between abstract and concrete developmental idealism (Allendorf, 2017). Abstract developmental idealism refers to schemas that pertain to the nature of the world in general, a country, or other broader collectives. The belief that arranged marriage is declining over time in Nepal or that it is, in general, better for young people to choose their own spouses are examples of abstract developmental idealism. Concrete developmental idealism refers to individuals using developmental idealism as a guide for their own lives or that of their children. A parent’s (negative) valuation of approving their own child’s spouse is an example of concrete developmental idealism; a parent who does not value approving their child’s spouse is implicitly using developmental idealism as a guide for their own life and that of their child.

Hypothesis 1: Parents who endorse abstract developmental idealism about arranged marriage and parental respect are less likely to value approving their child’s future spouse (a form of concrete developmental idealism).

Before turning to other factors, it is important to note what this theory does not do. Developmental idealism theory and accompanying empirical research does not take a position on the veracity and desirability of the content of developmental idealism. The theory contends only that developmental idealism is widely adopted and influential—not that the values and beliefs found within it are actually correct or good.

Parents’ Marital Experience

Past research demonstrates that parents’ marital experiences influence children’s marital experiences (Sweeney, 2002; Thornton, 1991; Uecker & Stokes, 2008). Americans are more likely to marry at young ages if their parents married young. And, Nepali men progress into marriage sooner if their fathers’ had greater choice over their own marriages (Jennings, Axinn, & Ghimire, 2012). Intergenerational influences work through ideational and socioeconomic pathways. Socioeconomically, children’s education and economic circumstances often resemble those of their parents. Ideationally, parents’ typically value their own experiences and wish for their children to have similar experiences.

Hypothesis 2: Parents who chose their own spouse are less likely to value approving their child’s spouse than parents who had an arranged marriage.

Children’s Independence and Parents’ Strategic Tradeoffs

As noted above, convergence theory suggests that children’s economic independence influences parents’ valuation of spouse approval (Burgess & Locke, 1960; Goode, 1963). The wide variety of family behaviors found across the globe converge toward the Western family form due to the forces of industrialization and urbanization. As young people take up jobs in industrialized, urbanized economies, they become able to independently support themselves and, in turn, gain the power to choose their own spouses. Empirical studies support this theorized link; young Nepalis with nonfamily employment were more likely to choose their spouse than those who worked for their families or did not work (Allendorf & Thornton, 2015; Ghimire et al., 2006). Ethnographic studies further suggest that parents’ opting to step back is a pathway linking children’s economic independence to self-choice marriages (Allendorf, 2013; Corwin, 1977; Donner, 2002; Ross, 1961). Most evocatively, Corwin (1977, p. 830) identified children’s college education and employment as “psychological weapon[s]” against parents because they allow children to set up separate households.

Hypothesis 3: Parents are less likely to value spouse approval when their child is highly educated and employed (or employable) outside the family.

Some parents may also make strategic tradeoffs to avoid adverse consequences of children’s (potential) independence in the future. Specifically, parents may relinquish approval to maintain access to old age support. In South Asia, sons are customarily responsible for caring for parents in old age (Karve, 1965; Lamb, 2005). However, some sons do not fulfill this responsibility and the risk that sons will not provide support is a concern for parents. Allowing a son to choose his own wife, even contrary to the parent’s own preferences, may be an acceptable tradeoff to maintain a good relationship with their son and, in turn, access to future support. In Chitwan Valley, Nepal, several residents explicitly said that parents allow sons greater choice over their wives for this reason (C. Mitchell, personal communication, 2015).

Hypothesis 4: Parents are less likely to value spouse approval when they are concerned about securing old age support.

Relationship Quality

Parents may also be willing to relinquish approval to maintain good relations and proximity to their children for its own sake. When children marry without parents’ approval, they may stop coresiding with their parents or even break ties entirely with their family (Allendorf, 2013; Chowdhry, 2007; Mody, 2008). Some parents may happily give up approval to avoid such a fate. For example, in Corwin’s (1977) study, a father reported giving his daughter consent to marry a man that he did not approve of because he feared she would go to Calcutta and marry there without his consent if he did not. This case suggests parents with good relationships with children may be more willing to give up approval in order to please children and maintain relationships.

At the same time though, parents with strong relationships with children may be less willing to give up approval. While parents’ approval of a child’s spouse is often conceptualized as an aspect of parental control, it is also a manifestation of parental concern. The choice of a spouse has profound consequences for a child’s future. Parents with high-quality relationships with their children may highly value approval because they want to ensure a beloved child marries well. Conversely, parents with poor-quality relationships may be less concerned about their child’s welfare and are content to let their child make their own choice, for good or ill. When they have poor relationships, parents may also not value approval because they adjust to the reality that they are unlikely to be able to influence their child’s choice of a spouse.

Hypothesis 5: Parents with low-quality relationships with their child are less likely than those with high-quality relationships to value spouse approval.

Gender Differences

The hypotheses described above were framed largely in gender neutral terms, applying to mothers and fathers in relation to sons and daughters. However, Nepal, like much of South Asia, is a highly gender stratified society. Roles and responsibilities of mothers and fathers, as well as sons and daughters, differ in profound ways. Many of these differences are tied to the joint family system, which is patrilocal and patrilineal (Bennett, 1983; Karve, 1965). Customarily, sons inherit family property, reside with parents throughout their lives, and are responsible for caring for parents in old age (Das Gupta, 1999). Conversely, daughters join their husband’s family after marriage, do not inherit natal family property, reside with in-laws after marriage, and care for elderly parents-in-law, rather than their own parents. Furthermore, men usually have greater decision-making power than women, higher levels of education, are more likely to own assets, and are more likely to work outside the family and earn higher incomes when they do (Acharya & Bennett, 1981; Allendorf, 2007). Marriage is also tightly linked to women’s, but not men’s, sexuality. Women are expected to maintain virginal reputations until marriage and are often expected to behave modestly, obey parents, and refrain from mixing with men to keep their reputations intact (Abraham, 2001; Brunson, 2014; Desai & Andrist, 2010).

Given this context, approving a child’s spouse may be more important to mothers and in reference to daughters. I expect that concerns about maintaining daughters’ reputations, valuation of daughters’ obedience, and daughters’ dependency on future husbands motivate parents to value approving a daughter’s husband more than a son’s wife (cf. Allendorf, 2017). I also expect that mothers value approval more than fathers because such valuation is consistent with women’s customary role as caretaker of the family. Since marriage is more salient for women’s identity and welfare, mothers may carry over the importance of marriage in their own lives to concerns about their children’s marriage, whether the child is a son or daughter. Fathers may be less tied to a family caretaker role and less directly concerned about children’s marriages.

When turning to children’s gender, I expect two factors to be more important for parents of sons, while a third should be more important for parents of daughters. First, making strategic tradeoffs in order to ensure old age support is pertinent to sons because it is sons who are expected to provide old age support. I also expect sons’ potential independence, particularly their employment, to be more influential than daughters’ potential independence. Sons are expected to contribute to the family livelihood, while daughters are not. So, I expect that sons’ employment status will have more influence on parents. Conversely, I expect that quality of relationships with daughters will be more influential than those with sons. Even if parents have poor relationships with sons, they may still value spouse approval because his wife is expected join their family. Conversely, sons-in-law remain primarily members of their own families.

When considering the gender of parents themselves, I expect that strategic tradeoffs are more influential for mothers and marital experience are more influential when parents’ and children’s genders match. Since women are less likely to be financially independent, mothers may be more concerned than fathers with maintaining access to old age support. Furthermore, since husbands are usually older than wives, most women will spend time as widows. In turn, I expect mothers are more concerned about old age support than fathers. I also expect that mothers see their marital experience as more relevant to daughters, while fathers view their experience as more relevant to sons.

Method

Data

The data were collected in 2008-2012 under the Chitwan Family Valley Study. Chitwan Valley is located in south central Nepal, just north of the Indian border. Data collection began in 2008 with a baseline survey of all individuals aged 15 to 59 years that were members of households located within sample neighborhoods, as well as spouses and some parents of the individuals living in selected households. Members of households that were 12 to 14 years old in 2008 were also administered the survey, but only after they turned 15 years old. This younger cohort was interviewed on an ongoing basis from 2008 to 2012. The final sample included 5,802 individuals aged 15 to 59 years at the time of interview and had a response rate of 97.1%.

The target population is parents for whom approving a child’s future spouse is relevant. In turn, the analytical sample comprises parents of young people at risk of first marriage, which I define as never married youth aged 15 to 24 years at baseline. Some parents of these youth were included in the survey as members of the original target population because they coresided with their children in selected households and were less than 59 years old. However, the survey was designed to capture all parents of such never married youth; parents of these youth were surveyed as long as they were alive and lived in Nepal. In total, there were 1,768 youth at risk of first marriage and only 206 had one or both parents missing from the survey.

To ensure that all children in the analytical sample had data from their mother and father, the analytical sample is anchored on the 1,562 never married youth aged 15 to 24 years at the time of interview who had both parents included in the survey. (These youth include 972 aged 15 to 24 years in 2008 and 590 aged 12 to 14 years in 2008.) Some measures used in the analysis pertain to the child, others to the parent, and a few to the parent–child pair. So the analytical sample can be conceptualized as 3,124 parent–child pairs or 1,562 mother–father–child trios, including 799 mother–father–son trios and 763 mother–father–daughter trios.

Measures

Dependent Variable

Parents’ valuation of spouse approval is measured with the following question: “How important is it to you that [child’s name] marry a person you approve of/choose? Would you say very important, somewhat important, or not important at all?” All three response options are retained. Crucially, this question concretely refers to the parents own, specific child by name and is unique to the parent–child pair. The measure does not assess if parents values spouse approval in general or abstract terms. It is also important to emphasize that, as the English translation implies, the original Nepali phrasing encompasses choosing or approving a child’s spouse. A parent valuing choosing the child’s spouse implies the parent wants an arranged marriage for their child. However, valuing approving the spouse is also consistent with preferring a hybrid marriage in which the child may identify the prospective spouse, but the marriage is contingent on the parent approving that choice.

Independent Variables

Five variables assess parents’ endorsement of abstract developmental idealism about arranged marriage and parental respect. Two of these variables assess parents’ beliefs about future changes. Specifically, whether parents said that (1) respect for parents and (2) arranged marriages will decrease in the future in Nepal. Two other measures assess parents’ valuation of possible changes. Namely, whether parents said it would be good if (3) respect for parents and (4) parents choosing spouses decrease in the future in Nepal. Another measure assesses parents’ valuation of arranged marriages without reference to change over time. Namely, whether it is (5) better for children to choose their own spouse, rather than parents choosing children’s spouses. All five of these measures are abstract because they measure parents’ values and beliefs about spouse choice and intergenerational relations in reference to Nepal as a whole or for people in general.

Parents’ own marital experience is measured with a question about who chose their (first) spouse: “People marry in different ways. Sometimes our parents or relatives decide whom we should marry, and sometimes we decide ourselves. In your case, who selected your (first) spouse? Your parents/relatives, yourself, or both?” All three response options were retained.

Children’s (potential) independence is measured with education and nonfamily work experience at the time of interview. The work measure denotes whether the child ever worked outside the family and, if so, whether they ever held a salaried job. Child’s education is measured with a categorical variable that denotes whether they had 0 to 6, 7 to 9, 10 to 11, or 12+ years of education. Education is a marker of status, as well as an important asset in labor and marriage markets.

Strategic tradeoffs is measured with parents’ agreement to the following statement, “Sons who are not allowed to choose their own spouses may move away from their parents’ homes and not look after them later in life.” It is important to note that this measure is abstract—it does not assess whether the parent is willing to make such a tradeoff with their own son or concerned about securing old age support in the future. Instead, it assesses whether the parent believes that sons abandoning elder care responsibilities is, in general, a consequence of not being allowed to choose their wives.

Quality of the parent–child relationship is measured by the parents’ assessment of how happy their relationship with this particular child is. Parents were asked, “In general, would you say that your relationship with [child’s name] is extremely happy, very happy, somewhat happy, or not happy at all?” Only a handful of parents said that the relationship was not happy at all. These parents were collapsed into the “somewhat happy” category, making “somewhat happy” the lowest quality category.

Controls

The analysis includes control variables for several characteristics that are correlated with parents’ valuation of spouse approval and the independent variables. These controls include parents’ education, parents’ age, children’s age, the household’s distance from the urban center, and caste/ethnicity. The caste/ethnicity measure denotes four major caste/ethnic categories in Nepal, including Chhetri-Bahun (high castes), Dalits (low castes), Hill Janajati, and Terai Janajati.

Descriptive statistics for variables that pertain to the parent or parent–child pair are shown in Table 1 and presented separately for (1) mothers of sons, (2) mothers of daughters, (3) fathers of sons, and (4) fathers of daughters. Descriptive statistics for variables that pertain to the child or household are shown in Table 2 and presented separately for sons and daughters.

Table 1.

Descriptive Statistics for Parents by Gender of the Parent and Their Child.

					Distributions differ by parent’s gender?		Distributions differ by child’s gender?
	Mothers of sons (n = 799), %	Mothers of daughters (n = 763), %	Fathers of sons (n = 799), %	Fathers of daughters (n = 763), %	Among parents of sons	Among parents of daughters	Among mothers	Among fathers
	(1)	(2)	(3)	(4)	(1 vs. 3)	(2 vs. 4)	(1 vs. 2)	(3 vs. 4)
Importance of approving this child’s spouse					*	*	**	**
Very important	37.7	43.3	31.8	36.6
Somewhat important	41.9	42.2	43.7	46.0
Not important	20.4	14.6	24.5	17.4
Respect for parents will decrease in future	49.4	47.6	45.4	41.9	ns	*	ns	ns
Arranged marriages will decrease in future	83.7	83.9	89.4	88.3	**	*	ns	ns
Good if respect for parents decreases	6.6	7.3	7.3	7.1	ns	ns	ns	ns
Good if parents choosing spouses decreases	35.7	41.6	38.9	40.2	ns	ns	*	ns
Better for children to choose own spouse	69.0	69.2	79.7	81.8	***	***	ns	ns
Spouse choice experience					***	***	ns	ns
Parents and/or relatives	73.2	73.7	44.4	46.0
Both myself and parents	6.5	8.1	19.3	18.6
Myself	20.3	18.2	36.3	35.4
Sons may leave if don’t choose wife	41.4	35.5	39.2	38.9	ns	ns	*	ns
Happiness of relationship with this child					*	ns	ns	***
Extremely happy	53.9	54.0	50.4	54.9
Very happy	38.8	38.8	38.2	40.2
Somewhat happy	7.3	7.2	11.4	4.9
Education					***	***	ns	*
None	61.7	62.4	19.9	25.4
1-9 Years	30.2	29.9	54.7	49.9
10+ Years	8.1	7.7	25.4	24.6
Age					***	***	ns	ns
<40 Years	54.2	49.5	27.3	23.6
40-49 Years	35.8	37.4	48.6	49.3
50+ Years	10.0	13.1	24.2	27.1

Note. ns = nonsignificant.

p < .05. **p < .01. ***p <.001 (chi-square tests).

Table 2.

Descriptive Statistics for Children and Households by Gender of the Child.

	Sons (n = 799), %	Daughters (n = 763), %	Differs by child’s gender?
Child’s education			*
0-6 Years	19.9	14.8
7-9 Years	52.2	54.8
10-11 Years	17.9	21.0
12+ Years	10.0	9.4
Child’s nonfamily work experience			***
None	68.0	55.6
Unsalaried work	22.7	40.5
Salaried work	9.3	3.9
Child’s age, M (SD)	16.2 (3.2)	16.2 (3.1)	ns
Caste/ethnicity			ns
Chhetri-Bahun	45.6	47.2
Dalit	12.3	9.3
Hill Janajati	22.0	19.8
Terai Janajati	20.2	23.7
Distance from urban area, M (SD)	8.3 (4.0)	8.6 (4.0)	ns

Note. ns = nonsignificant. The “differ?” column reports the significance level of chi-square and two-tailed t tests of whether the distribution of the variable differs by the child’s gender.

p < .05. **p < .01. ***p < .001.

Analytical Strategy

First, I examined variation in parents’ valuation of spouse approval and evaluate whether a sizeable proportion of parents do not value approving their child’s future spouse. Next, I turned to the central research question and examine which parents value approval less than others. This analysis is based on a multinomial logit model.

On its face, the dependent variable, parents’ valuation of spouse approval, is ordinal. So an ordered logistic regression model may have been a better choice. However, ordered logistic regression models failed to uphold the required parallel regression assumption, in which the odds of being in any higher category versus a lower category are assumed to be equal. An alternative is the generalized ordered logit model, which enforces the parallel regression assumption for variables it applies to and relaxes it for the rest (Williams, 2006). I chose a multinomial model over a generalized ordered logit because the dependent variable may not be fully ordinal. Rather than having an intermediate or middling valuation of approval, parents in the middle category of “somewhat important” may feel the contradictions of simultaneously strongly valuing and not valuing approving their child’s spouse. The multinomial model has the advantage of showing how each independent variable is associated with all three spouse approval categories, rather than just showing the extent to which it is associated with less (or more) valuation of approval.

Results are drawn from only one, full model that includes all independent variables and controls, as well as three-way interactions of each of the independent and control variables by the parent’s and child’s gender. These interactions allow the effects of each independent variable to vary by the gender of the parent and the child. Thus, this approach is equivalent to running four separate models, one each for (1) mothers of sons, (2) mothers of daughters, (3) fathers of sons, and (4) fathers of daughters. The model is adjusted for clustering within neighborhoods via fixed effects and for clustering within households using robust standard errors.

I present results as average marginal effects for each parent–child gender pairing. I used the delta method to test whether the marginal effects differ from zero and differ by parents’ and children’s gender (Long & Freese, 2014). Average marginal effects are differences in the predicted probabilities of parents valuing spouse approval between the reference category of an independent variable versus other category(ies) of the independent variable when all other variables are held at observed values. For example, an average marginal effect for child’s salaried work of −.20 for approval is very important would indicate that the predicted probability of a parent of a child with salaried work valuing approval as very important differs by −.20 from the predicted probability for the reference category in which the parents’ child has no work experience. The marginal effects must sum to zero across the three categories of spouse approval. So, if the average marginal effect of salaried work on saying approval is very important is −.20, the average marginal effects for the somewhat important and not important categories may be .15 and .05, respectively (−.20 + .15 + .05 = 0). Such a result would indicate that parents whose children had salaried work were more likely to value spouse approval as somewhat important and, to a lesser extent, not important, compared with parents in the reference category whose children have no work experience.

One limitation of the analysis is the cross-sectional research design. Any associations between parents’ valuation of spouse approval and other ideational measures may be due to reverse causation. The theory of reasoned action and empirical evaluations of that theory indicate that the bulk of causal influences do move from more general or abstract schemas to more concrete and specific schemas (Fishbein & Ajzen, 2010). This work supports my conceptualization, but influences may still be moving in the reverse direction. This limitation does not apply to the nonideational variables though. The measures of children’s education, children’s work experience, parents’ marital experience, and parents’ education all refer to characteristics that were set prior to the time of the survey.

Results

Parents do show meaningful variation in valuing spouse approval. A sizeable proportion of parents, 19%, said approving their child’s spouse is not important (Figure 1). Nearly twice as many parents, 37%, were at the opposite end of the spectrum, saying such approval is very important. The modal category, by a modest percentage, is the middle category; 43% of parents said approving their child’s spouse is somewhat important.

Parents were more likely to value approving daughters’ husbands than sons’ wives. These differences were statistically significant, but modest in size. 40% of parents said approving their daughter’s future husband is very important, while 35% said the same in relation to son’s future wives (Figure 1). At the other end of the spectrum, 16% of parents of daughters said approval is not important, while 23% of parents of sons did so. The remainder, 44% and 43% of parents of daughters and sons, respectively, said approval is somewhat important.

Mothers were more likely than fathers to value approval. These differences by parents’ gender were also statistically significant and modest in size. 40% of mothers said approval is very important versus 34% of fathers (Figure 1). At the other end of the spectrum, 18% of mothers said approval is not important, while 21% of fathers did so. And, in the middle, 42% of mothers and 45% of fathers said it is somewhat important. (More detailed breakdowns by gender of the parent and child simultaneously are presented in Table 1.)

Next, I describe the results of the multinomial logit model that was used to test the hypotheses given above. Average marginal effects for parents of daughters are presented in Table 3, while those for parents of sons are in Table 4.

Table 3.

Average Marginal Effects From the Multinomial Logit Model of the Importance of Parents Approving Their Daughter’s Future Husband.

	Mothers			Fathers
	Approval is very important	Approval is somewhat important	Approval is not important	Approval is very important	Approval is somewhat important	Approval is not important
Respect for parents will decrease^a	.06	−.03	−.03	−.09**	.10**	−.02
Arranged marriages will decrease	−.01	−.02	.04	−.09	−.01	.10*
Good if respect for parents decreases^b	.08	−.06	−.02	−.10	.01	.09
Good if parents choosing spouses decreases^a	−.09*	.11**	−.02	−.04	−.04	.07
Better for children to choose spouse^c	−.13**	.04	.09**	−.19***	.05	.14***
Spouse choice experience
Parents and/or relatives (ref)	0	0	0	0	0	0
Both myself and parents	−.05	−.04	.09	−.08	.08	.00
Myself	.02	−.01	−.01	.00	−.02	.01
Child’s education
0-6 Years (ref)	0	0	0	0	0	0
7-9 Years^a	.10*	−.14**	.04	.00	.03	−.03
10-11 Years	.07	−.12	.05	−.00	.02	−.02
12+ Years	.05	−.06	.01	−.02	.02	−.00
Child’s nonfamily work experience
None (ref)	0	0	0	0	0	0
Unsalaried work	.01	−.06	.05	−.03	.03	.00
Salaried work^c	−.26***	.14	.13	−.13	.04	.10
Sons may leave if don’t choose wife	.01	.01	−.02	.01	.06	−.07*
Happiness of relationship with child
Extremely happy (ref)	0	0	0	0	0	0
Very happy	−.00	−.02	.02	−.04	.04	−.01
Somewhat happy	−.22***	.24***	−.02	−.10	.15	−.05

Note. The model also includes controls for parents’ education, parents’ age, child’s age, caste, and distance from the urban area.

The difference between mothers’ (of daughters) marginal effects and fathers’ (of daughters) marginal effects are statistically significant (p < .05; i.e., gender of parents is statistically significant for daughters). ^bThe difference between fathers’ marginal effects on sons and fathers’ marginal effects on daughters is statistically significant (p < .05; i.e., gender of child is statistically significant for fathers). ^cThe difference between mothers’ marginal effects on sons and mothers’ marginal effects on daughters is statistically significant (p < .05; i.e., gender of child is statistically significant for mothers).

p < .05. **p < .01. ***p < .001.

Table 4.

Average Marginal Effects From the Multinomial Logit Model of the Importance of Parents Approving Their Son’s Future Wife.

	Mothers			Fathers
	Approval is very important	Approval is somewhat important	Approval is not important	Approval is very important	Approval is somewhat important	Approval is not important
Respect for parents will decrease	.02	−.02	.00	−.03	.01	.01
Arranged marriages will decrease	−.09	.05	.04	−.01	−.09	.10*
Good if respect for parents decreases^a	.01	.12	−.13**	.06	.03	−.09
Good if parents choosing spouses decreases^b	−.04	.07	−.03	−.09*	−.02	.12**
Better for children to choose spouse^c	−.16***	−.00	.16***	−.25***	.05	.20***
Spouse choice experience
Parents and/or relatives (ref)	0	0	0	0	0	0
Both myself and parents	.04	−.02	−.02	−.07	.02	.05
Myself	−.05	.02	.03	−.04	.02	.02
Child’s education
0-6 Years (ref)	0	0	0	0	0	0
7-9 Years	−.03	−.01	.04	−.05	.07	−.02
10-11 Years	−.02	.01	.01	−.02	−.03	.06
12+ Years	.01	−.07	.06	.05	−.01	−.04
Child’s nonfamily work experience
None (ref)	0	0	0	0	0	0
Unsalaried work	−.02	.01	.01	.01	.02	−.03
Salaried work^c	−.07	.02	.05	.01	−.01	−.00
Sons may leave if don’t choose wife	−.02	.01	.01	.03	.01	−.05
Happiness of relationship with child
Extremely happy (ref)	0	0	0	0	0	0
Very happy	−.01	.01	.00	−.02	.04	−.02
Somewhat happy	−.14*	.06	.08	−.02	.04	−.02

Note. The model also includes controls for parents’ education, parents’ age, child’s age, caste, and distance from the urban area.

The difference between fathers’ marginal effects on sons and fathers’ marginal effects on daughters is statistically significant (p < .05; i.e., gender of child is statistically significant for fathers). ^bThe difference between mothers’ (of sons) marginal effects and fathers’ (of sons) marginal effects are statistically significant (p < .05; i.e., gender of parents is statistically significant for sons). ^cThe difference between mothers’ marginal effects on sons and mothers’ marginal effects on daughters is statistically significant (p < .05; i.e., gender of child is statistically significant for mothers).

p < .05. **p < .01. ***p < .001.

Parents’ endorsement of abstract developmental idealism about spouse choice had large, statistically significant marginal effects. Parents who said it was better for children to choose their own spouse generally did not value spouse approval. For example, if a father said it was better for children to choose their own spouses, the predicted probability of him saying spouse approval is not important increased by .14 in relation to daughters and .20 in relation to sons (Tables 3 and 4). Conversely, fathers’ predicted probability of saying approval is very important was reduced by .19 for daughters and .25 for sons if the father said it was better for children to choose their own spouses. Similar, but more modest, results are seen for mothers; mothers’ marginal effects are .09 and .16 for saying approval is not important and −.13 and −.16 for very important in relation to daughters and sons, respectively (Tables 3 and 4).

The other development idealist value about change over time in spouse choice shows modest effects on parents’ valuation of approval. Although the pattern differs significantly by the gender of the parent and the marginal effects are not consistently statistically significant. Fathers who said it was good if parents choosing spouses decreases were less likely to value approval. Specifically, for fathers of daughters, the marginal effect for approval is not important is .07, while the marginal effects are −.04 and −.04 for very and somewhat important (Table 3). Similarly, for fathers of sons, the marginal effect for approval is not important is .12, while the marginal effects are −.09 and −.02 for very and somewhat important (Table 3). By contrast, mothers who value decreases in parents choosing spouses are more likely to be ambivalent about approving their child’s spouse; such mothers are more likely to say approval is somewhat important and less likely to say it is very or not important. For mothers of daughters, the marginal effect for approval is somewhat important is .11, while it is −.09 for very important and −.02 for not important (Table 3). Similarly, for mothers of sons, the marginal effect for somewhat important is .07, while it is −.04 for very important and −.03 for not important.

The third developmental idealist schema on spouse choice shows small marginal effects in the expected direction, but none are statistically significant. Parents are slightly less likely to value approval if they think arranged marriage will decrease. The marginal effects for saying spouse approval is not important range from .04 to .10, while those for very or somewhat important range from −.09 to −.01 (Tables 3 and 4). There is one exception to this pattern though. Mothers of sons have a marginal effect of .05 for the somewhat important category.

By contrast, the marginal effects for the developmental idealist schemas about parental respect do not have a consistent pattern, which suggests that these schemas are largely unrelated to parents’ valuation of approving their child’s spouse. For the belief that respect for parents will decrease over time, the marginal effects are small and often close to zero, ranging largely from −.03 to .06 (Tables 3 and 4). There is also not a consistent pattern of the somewhat or not important categories showing larger and more positive effects relative to the very important category. There is one exception to this pattern, however—for fathers of daughters the marginal effects for somewhat and very important are a statistically significant .10 and −.09, respectively (Table 3). This exception indicates that fathers who believe in a future decline in parental respect are moderately more likely to be ambivalent about approving their daughter’s future husband.

The marginal effects for the valuation of future declines in parental respect are generally larger in size, but many are not in the hypothesized direction and none are statistically significant. For example, in keeping with expectations, the marginal effect of .09 for the not important category indicates that fathers of daughters who said it is good if respect for parents declines were modestly less likely to value spouse approval (Table 3). Contrary to expectations, however, mothers of daughters who said it is good if parental respect declines were modestly more likely to value spouse approval as very important (marginal effect = .08, Table 3).

Parents’ own marital experiences are also not an important correlate of their valuation of approval. Marginal effects for parents’ own marital experience are near zero, ranging largely from −.05 to +.05, and none are statistically significant (Tables 3 and 4). Parents who chose their own spouses were, with one exception, slightly more likely to say approval is not important; the marginal effects for the not important category are .01, .03, and .02 for fathers of daughters, mothers of sons, and fathers of sons, respectively (Tables 3 and 4). These results are suggestive of potentially tiny influences of parents’ own marital experiences. It should be noted, however, that this pattern did not hold for mothers of daughters (Table 3). These null results also indicate that parents’ marital experiences are not more influential when the gender of the child and parent match.

Children’s potential independence shows mixed results. Contrary to expectations, parents were not consistently less likely to value approval when their child is highly educated. The marginal effects for children’s education are generally near zero and lack statistical significance, generally ranging from −.07 to .07 (Tables 3 and 4). There are three marginal effects that are relatively large in size at .10, −.14, and −.12 and two of them are statistically significant (Table 3). Crucially though, the pattern in the size of the effects is not consistent with high levels of children’s education reducing parents’ valuation of approval.

Children’s salaried work experience does influence parents’ valuation of approval. There are also seemingly large, but not always statistically significant, differences by the gender of the child and the parent with salaried work experience appearing more influential for mothers and parents of daughters. Most strikingly, the predicted probability of valuing approval as very important is .26 less among mothers of daughters with salaried jobs versus mothers of daughters who never worked (Table 3). This large reduction in mothers of daughters valuing approval as very important is balanced by increases in mothers of daughters with salaried jobs saying that approval is both somewhat important (marginal effect = .14) and not important (marginal effect = .13; Table 3). Fathers of daughters with salaried work also value approval less, but the point estimates are about half the size and not statistically significant (Table 3). The point estimates further suggest that sons’ salaried work experience may have similar, but substantially smaller impacts on mothers. Mothers of sons with salaried work experience were also less likely to value approval; the marginal effect for not important is .05, while the marginal effects for very and somewhat important are −.07 and .02 (Table 4). These smaller effects for mothers of sons do not differ significantly from zero, but they do differ significantly from the larger marginal effects for mothers of daughters. For fathers, however, sons’ salaried work experience is not important; the marginal effects are roughly zero at .01, −.01, and .00 (Table 4). Unsalaried work of both sons and daughters is not an important influence on parents though; the marginal effects are uniformly small, lack statistical significance, and do not show signs of a consistent relationship with parents’ valuation of approval.

Parents’ belief in strategic tradeoffs is also not an important correlate of valuation of approval. As expected, believing that sons may leave parents if the sons are not allowed to choose their own wives has no effect on mothers of daughters; the marginal effects are all close to zero, ranging from −.02 to .01 (Table 3). Oddly, there are indications of this belief having a modest effect on fathers of daughters, but it is not in the hypothesized direction; the marginal effect for not valuing approving is a significant −.07 for fathers of daughters who believe sons will leave if they do not choose their wife. Moreover, this belief had no effect on parents of sons, nor did it have a larger effect on mothers (Table 4).

Finally, the quality of the parent–child relationship is an important correlate of valuing approval; parents with low-quality relationships are more ambivalent about approval. Compared with parents with the best relationships with their children, those with the worst relationships were uniformly more likely to say approval is somewhat important and less likely to say approval is very important. The marginal effects on somewhat important for parents with somewhat happy relationships are .24, .15, .06, and .04, while those for very important are −.22, −.10, −.14, and −.02 (Tables 3 and 4). The largest of these marginal effects are also statistically significant, namely the two large effects of −.22 and .24 for mothers of daughters and the sizeable effects of −.14 for mothers of sons (Tables 3 and 4). The pattern in the size of the marginal effects further suggests that relationship quality is especially important for mothers and parents of daughters, and especially so when these genders overlap in the case of mothers of daughters. However, these differences by gender of the parent and gender of the child were not statistically significant. The marginal effects for mothers of daughters did differ significantly from those for fathers of sons though.

Discussion and Conclusion

In recent decades, as arranged marriage declined in Asia, we have made important strides in identifying which young people are more likely to choose their own spouses. A few intergenerational studies have also examined how parents’ characteristics influence their children’s marital behavior, including who chooses the spouse. In this article, I contribute to this work by giving further attention to parents. Rather than examining parents’ influence on children’s marriage behavior, this article takes a step back and focuses squarely on parents’ valuation of approving a child’s spouse before a marriage takes place.

There was substantial variation in parents’ valuation of spouse approval, with a sizeable proportion of parents saying approving their child’s spouse is not important. Specifically, a fifth of parents did not value approval, just over two fifths appeared ambivalent, and just under two fifths strongly valued it. Approval was also more valued by and about women; parents were more likely to value approving daughters’ husbands than sons’ wives and mothers were more likely than fathers to value approval. These gender differences were modest in size though.

Differences by children’s gender are consistent with compatibilities and contradictions between South Asian gender schemas and developmental idealist schemas (Allendorf, 2017). Using these same data, Allendorf (2017) similarly found that parents are more likely to expect their sons to choose their own spouse and do so for love. This gender difference in parents’ valuation of approval is also consistent with marriage behavior; men have greater control over the choice of spouse then women in many Asian contexts (Allendorf, 2017; Caldwell, Reddy, & Caldwell, 1983; Malhotra, 1991; Schuler, Bates, Islam, & Islam, 2006).

Some, but not all, aspects of developmental idealism were connected to parents’ valuation of approval. As hypothesized, parents who endorsed abstract developmental idealism about arranged marriage were less likely to value spouse approval. Specifically, parents who valued self-choice marriages as inherently better than arranged marriages were substantially less likely to value approving their own child’s spouse. Those who believed that arranged marriages will decrease over time also showed a tendency to value approval less, but the effects were small and not statistically significant.

Contrary to expectations, however, developmental idealism about parental respect was not an important correlate. Parents who said respect for parents will decrease over time or that such decreases are good did not systematically value approval less. This result suggests parents’ involvement in a child’s marriage plays a minor or nonexistent role in parents’ conceptualization of parental respect. The greater importance of developmental idealism about arranged marriage versus parental respect is consistent with prior studies, which found that schemas about the behavior in question are more influential than schemas about related behaviors (Allendorf, 2017; Allendorf & Thornton, 2015).

Some, but not all, aspects of children’s independence were important influences on parents’ valuation of approval. Specifically, children’s salaried work experience, particularly daughters’ work experience, was highly influential, while children’s unsalaried work experience and education were not. This gender difference ran counter to expectations; I hypothesized that sons’ work experience would be more influential than daughters’ experience.

This unexpected result may be due to multiple reasons. Parents of daughters with salaried work may value approving their son-in-law’s less because they view their daughters as especially able to encounter and select desirable husbands, which the parents may not have the same opportunities to find. Daughters’ with salaried work experience are rare and capable young women who are likely to meet men at work who are highly eligible on the marriage market. Furthermore, parents who enable daughters to obtain salaried work likely value independence for girls more broadly. Parents who are committed to daughters’ achievements and independence may allow their daughters to engage in salaried work and choose their own husband.

The other important correlate of parents’ valuation of approval is the quality of the relationship. Parents with low-quality relationships with their child were ambivalent about approval, saying they valued it somewhat, rather than very much or not at all. This result is consistent with parents being less concerned about children with whom they have poor quality relationships or simply adjusting to a reality in which they may not have much influence over their child’s marriage even if they wanted to. However, tipping into the intermediate category, rather than simply saying they do not value approval at all, suggests parents maintain some concern and/or valuation of controlling children’s marriages even when the relationship is poor.

The remaining factors, strategic tradeoffs and parents’ own spouse choice experience, did not emerge as important. Given changes in marriage over time, parents’ may view their own marital experiences as out of date and not salient to their children’s marriages. It is also possible that parents’ marital experiences do influence their valuation of approval, but the link is moderated by evaluation of their own marriages. A parent who wanted to choose their own spouse, but married a person their parents chose may not value approval. Instead, they may wish for their child to make their own choice, as they had wanted to. Parents’ valuation may also be tied to how their marriage turned out. For example, a parent who chose their own spouse, but then had a difficult marriage, may want to protect their child from making a similar mistake and may strongly value approval as a way to ensure their child marries well.

It is surprising that beliefs about strategic tradeoffs had no effect, given that individuals living in Chitwan, the survey site, said that this was an important reason why some parents allow sons to choose their own wives in in-depth interviews. This null result may be due to the nature of the measure; as noted above, it is an abstract measure of whether parents believe in general that parents lose old age support from their sons if the sons are not given the freedom to choose their own wives. It seems likely though that there is a strong correlation between concrete concern about one’s own situation and endorsement of the abstract belief about what happens in general. If so, strategic tradeoffs may simply not be an important consideration for parents. Parents’ need for old age support is distant in time and when it comes to their own children, parents may be concerned with more immediate issues or believe that any breech in the relationship over a son’s marriage would heal well before they need old age support.

Overall, parents living in a context of marital change were not monolithically in favor of maintaining control over, or even participation in, the choice of their child’s spouses. There was a sizeable reserve of parents who likely facilitate self-choice marriages of their children, by standing aside from their customary role of marriage arranger and eschewing even a more hybrid role of approving their child’s choice. Of course, parents’ behavior at the time of the child’s marriage, may not always match their valuation of approval prior to the marriage. Some parents may end up approving a child’s spouse even when they do not value it ahead of time and others may not approve a child’s spouse even when they do strongly value it. The other end of this spectrum should also be emphasized though. Even in this context in which self-choice marriage was comparatively common during the study period, substantial numbers of parents still valued approving their child’s spouse.

Footnotes

Acknowledgements

The author would like to thank members of the Developmental Idealism Working Group for their helpful comments.

Author’s Note

An earlier version of this article was presented at the Annual Meeting of the American Sociological Association in Montreal on August 25, 2017.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Keera Allendorf

References

Abraham

(2001). Redrawing the Lakshman Rekha: Gender differences and cultural constructions in youth sexuality in urban India. South Asia: Journal of South Asian Studies, 24, 133-156.

Acharya

Bennett

(1981). The status of women in Nepal. Kathmandu, Nepal: Tribhuvan University Centre for Economic Development and Administration.

Ahearn

L. A.

(2001). Invitations to love: Literacy, love letters and social change in Nepal. Ann Arbor: University of Michigan Press.

Allendorf

(2007). Couples’ reports of women’s autonomy and health-care use in Nepal. Studies in Family Planning, 38, 35-46. doi:10.1111/j.1728-4465.2007.00114.x

Allendorf

(2013). Schemas of marital change: From arranged marriages to eloping for love. Journal of Marriage and Family, 75, 453-469. doi:10.1111/jomf.12003

Allendorf

(2017). Conflict and compatibility? Developmental idealism and gendered differences in marital choice. Journal of Marriage and Family, 79, 337-355. doi:10.1111/jomf.12339

Allendorf

Pandian

R. K.

(2016). The decline of arranged marriage? Marital change and continuity in India. Population and Development Review, 42, 435-464.

Allendorf

Thornton

(2015). Caste and choice: The influence of developmental idealism on marriage behavior. American Journal of Sociology, 121, 243-287. doi:10.1086/681968

Axinn

W. G.

Ghimire

D. J.

Barber

J. S.

(2008). The influence of ideational dimensions of social change on family formation in Nepal. In Jayakody

Thornton

Axinn

W. G.

(Eds.), International family change: Ideational perspectives (pp. 251-280). Mahwah, NJ: Lawrence Erlbaum.

10.

Bennett

(1983). Dangerous wives and sacred sisters: Social and symbolic roles of high-caste women in Nepal. New York, NY: Columbia University Press.

11.

Brunson

(2014). “Scooty girls”: Mobility and intimacy at the margins of Kathmandu. Ethnos, 79, 610-629.

12.

Burgess

E. W.

Locke

H. J.

(1960). The family: From institution to companionship. New York, NY: American Book Company.

13.

Caldwell

J. C.

Reddy

P. H.

Caldwell

(1983). The causes of marriage change in South India. Population Studies: A Journal of Demography, 37, 343-361. doi:10.2307/2174503

14.

Chowdhry

(2007). Contentious marriages, eloping couples: Gender, caste, and patriarchy in Northern India. New Delhi, India: Oxford University Press.

15.

Cole

Thomas

L. M.

(2009). Love in Africa. Chicago, IL: University of Chicago Press.

16.

Corwin

L. A.

(1977). Caste, class and the love-marriage: Social change in India. Journal of Marriage and the Family, 39, 823-831.

17.

Das Gupta

. (1999). Lifeboat versus corporate ethic: Social and demographic implications of stem and joint families. Social Science & Medicine, 49, 173-184.

18.

Desai

Andrist

(2010). Gender scripts and age at marriage in India. Demography, 47, 667-687. doi:10.1353/dem.0.0118

19.

Donner

(2002). One’s own marriage: Love marriages in a Calcutta neighborhood. South Asia Research, 22, 79-94.

20.

Fishbein

Ajzen

(2010). Predicting and changing behavior: The reasoned action approach. New York, NY: Psychology Press.

21.

Ghimire

D. J.

Axinn

W. G.

Yabiku

S. T.

Thornton

(2006). Social change, premarital nonfamily experience, and spouse choice in an arranged marriage society. American Journal of Sociology, 111, 1181-1218. doi:10.1086/498468

22.

Goode

W. W.

(1963). World revolution and family patterns. London, England: Free Press.

23.

Harkness

Khaled

(2014). Modern traditionalism: Consanguineous marriage in Qatar. Journal of Marriage and Family, 76, 587-603.

24.

Hirsch

J. S.

Wardlow

(2006). Modern love: The anthropology of romantic courtship and companionate marriage. Ann Arbor: University of Michigan Press.

25.

Jennings

E. A.

Axinn

W. G.

Ghimire

D. J.

(2012). The effect of parents’ attitudes on sons’ marriage timing. American Sociological Review, 77, 923-945. doi:10.1177/0003122412464041

26.

Karve

(1965). Kinship organization in India. Bombay, Maharashtra, India: Asia Publishing House.

27.

Kaur

Palriwala

(Eds.). (2014). Marrying in South Asia: Shifting concepts, changing practices in a globalizing world. Hyderabad, Telangana, India: Orient Black Swan.

28.

Lamb

(2005). Cultural and moral values surrounding care and (in)dependence in late life: Reflections from India in an era of global modernity. Care Management Journals, 6, 80-89.

29.

Long

J. S.

Freese

(2014). Regression models for categorical dependent variables using Stata, third edition. College Station, TX: Stata Press.

30.

Malhotra

(1991). Gender and changing generational relations: Spouse choice in Indonesia. Demography, 28, 549-570. doi:10.2307/2061422

31.

Marsden

(2007). Love and elopement in northern Pakistan. Journal of the Royal Anthropological Institute, 13, 91-108.

32.

Meekers

(1995). Freedom of partner choice in Togo. Journal of Comparative Family Studies, 26, 163-178.

33.

Mody

(2008). The intimate state: Love-marriage and the law in Delhi. Delhi, India: Routledge India.

34.

Nedoluzhko

Agadjanian

(2015). Between tradition and modernity: Marriage dynamics in Kyrgyzstan. Demography, 52, 861-882.

35.

Netting

N. S.

(2009). Marital ideoscapes in 21st-century India: Creative combinations of love and responsibility. Journal of Family Issues, 20, 1-20.

36.

Rebhun

L. A.

(1999). The heart is unknown country: Love in the changing economy of northeast Brazil. Stanford, CA: Stanford University Press.

37.

Rocca

C. H.

Rathod

Falle

Pande

R. P.

Krishnan

(2009). Challenging assumptions about women’s empowerment: Social and economic resources and domestic violence among young married women in urban south India. International Journal of Epidemiology, 38, 577-585. doi:10.1093/ije/dyn226

38.

Ross

A. D.

(1961). The Hindu family in its urban setting. Toronto, Ontario, Canada: University of Toronto Press.

39.

Schuler

S. R.

Bates

L. M.

Islam

M. K.

(2006). The timing of marriage and childbearing among rural families in Bangladesh: Choosing between competing risks. Social Science & Medicine, 62, 2826-2837.

40.

Sweeney

M. M.

(2002). Two decades of family change: The shifting economic foundations of marriage. American Sociological Review, 67, 132-147. doi:10.2307/3088937

41.

Thornton

(1991). Influence of the marital history of parents on the marital and cohabitational experiences of children. American Journal of Sociology, 96, 868-894.

42.

Thornton

(2001). The developmental paradigm, reading history sideways, and family change. Demography, 38, 449-465. doi:10.2307/3088311

43.

Thornton

(2005). Reading history sideways: The fallacy and enduring impact of the developmental paradigm on family life. Chicago, IL: University of Chicago Press.

44.

Uecker

J. E.

Stokes

C. E.

(2008). Early marriage in the United States. Journal of Marriage and Family, 70, 835-846.

45.

Williams

(2006). Generalized ordered logit/partial proportional odds models for ordinal dependent variables. Stata Journal, 6, 58-82.