Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Jun 1.
Published in final edited form as: J Marriage Fam. 2014 Apr 25;76(3):476–488. doi: 10.1111/jomf.12104

Marital Discord and Subsequent Marital Dissolution: Perceptions of Nepalese Wives and Husbands

Elyse Jennings 1
PMCID: PMC4251804  NIHMSID: NIHMS600625  PMID: 25484450

Abstract

This study examines the influence of marital discord on separation and divorce in a rural South Asian setting. We know little about how marital discord influences marital outcomes in settings with low personal freedom and limited access to independence. Using a sample of 674 couples from the Chitwan Valley Family Study in Nepal, this paper investigates the impact of marital discord on the rate of marital dissolution, and the extent to which wives’ and husbands’ perceptions of discord influence dissolution. Results reveal that (a) spouses’ perceptions of marital discord increase the rate of marital dissolution, (b) both husbands’ and wives’ perceptions of discord have an important influence, and (c) the influence of wives’ perceptions of discord is independent of their husbands’ perceptions. Overall, these findings suggest that both spouses’ perceptions of discord are important for marital outcomes, even in settings where the costs of marital dissolution are relatively high.

Keywords: Family, marital dissolution, marital quality, marriage, non-U.S. families

Over the last several decades populations outside of industrialized, Western settings have begun to adopt a new model of marriage that emphasizes personal fulfillment and happiness (Dion & Dion, 1993; Goode, 1970; Thornton, 2001). Accompanying this marital shift, many of these non-Western countries have seen a rising prevalence of divorce (Goode, 1993). Yet, we have a limited understanding of the process among these more collectivist-oriented populations. This study investigates the process of marital dissolution in rural South Asia: a setting where individuals, and particularly women, have relatively low personal freedom (Bennett 1983; Jayakody, Thornton, & Axinn, 2008; Sastry & Ross, 1998). Although research demonstrates that marital quality is associated with marital dissolution among Western populations (Amato & Rogers, 1997; DeMaris, 2000; Gottman, 1994), it is unclear whether marital quality will have a similar impact in this South Asian setting. It is even less clear whether a wife’s perception of marital quality will be as influential as her husband’s perception of marital quality.

Specifically, this paper focuses on the role of marital discord in predicting marital outcomes among an agrarian population in southern Nepal. Divorce remains uncommon throughout South Asia (Dommaraju & Jones, 2011), but is likely to be on the rise as families and their social surroundings are changing (Axinn & Yabiku, 2001; Jayakody et al., 2008). In fact, the number of divorces registered in the Chitwan District Court of southern Nepal has been increasing rapidly since the mid-1980s, from less than 10 per year, to over 200 in the year 2010. Furthermore, nearly every young Nepali person marries, which automatically exposes them to the possibility of experiencing marital dissolution. Hence, this setting is ideal for studying the association between marital discord and marital dissolution where divorce is becoming increasingly common.

BACKGROUND

In rural Nepal, the high value placed on marriage is indicated in its universality and early occurrence. The average age at marriage for those marrying between 2000 and 2005 was 19.9 for women and 23.9 for men (Yabiku, 2005). Most marriages have historically been arranged by family members (Ghimire, Axinn, Yabiku, & Thornton, 2006). The marital union is especially important for the gendered division of labor within households, as women are responsible for certain tasks both in the field and in the home (Allendorf, 2007). Women typically do not hold jobs outside of the home, although it is not uncommon for women to perform temporary work for wages. This temporary work often involves labor on the land of wealthier households and is somewhat deprecated (Cameron, 1998; Stash & Hannum, 2001). Because the women who work for wages tend to occupy a lower social status, they also tend to be under less strict expectations—compared to higher status women—to show obedience to their husbands and in-laws (Bennett, 1983).

It is not surprising, in a setting where marriage is nearly universal and encompasses well-defined norms and customs, that marital dissolution is uncommon (Parry, 2001). Hindu customs guide many aspects of social life in Nepal, and Hindu code historically has not formally recognized marital dissolution (Goode, 1970; Holden, 2008; Parry, 2001). Nevertheless, marital dissolutions are often informal, and many do not involve legal recognition (Parry, 2001). Lower status, marginalized groups have a history of greater instances of marital dissolution than higher status groups (Holden, 2008).

Data from in-depth interviews (conducted in Fall 2010 with 30 men and women of various Nepalese ethnicities who were living in Chitwan) reveal that local people tend to be aware that divorce occurs around them, but also view it as adverse or perceive that others view it as adverse. A 25-year-old woman said “When one gets married with a person he or she should … live without breaking their relation[ship]. Otherwise society looks with negative eyes.” As a result of this stigma, people can have a hard time finding a second spouse. For women, who have few opportunities to support themselves outside of marriage, this can present a particularly salient obstacle. Women may be left with no alternative than to return to their natal home, where they are sometimes not welcome. Separation can be a more desirable option for wives, as it can allow them to continue to receive support from their husbands.

Even so, the majority of marriages (63%) that dissolve among respondents of the Chitwan Valley Family Study (CVFS) are the result of divorce (without first separating). Separations can lead to divorce later on, but many separations last for long periods or indefinitely. Local people recognize that separation is a form of dissolution, like divorce, but unique in that it allows a wife to maintain economic support and allows the possibility of a reunion. A 27-year-old woman described the difference as follows: “When [a husband and wife] live separately they don’t have written document, they perform it orally due to quarrelling on a small matter. But divorce is a written document and states that they don’t have any relationship ….”

Nonetheless, marital dissolution is experienced by few members of the population. Among women in Chitwan, Nepal who first married in the 1980s, only 10% had experienced the dissolution of that marriage by 2008. This is much lower than a setting like the United States, for example, where recent estimates suggest that about half of marriages end in divorce (Cherlin, 2010).

CONCEPTUAL FRAMEWORK

Marital Discord in Nepal

Marital discord is a strong predictor of marital dissolution in Western settings, as both observational and survey methods of data collection have revealed (Amato & Rogers, 1997; DeMaris, 2000; Gottman, 1994; Matthews, Wickrama, & Conger, 1996). Theories based on a cost-benefit approach would predict that a marriage will dissolve once the benefits of leaving that marriage outweigh the costs (Becker, Landes, & Michael, 1977). This threshold may be relatively high in Nepal, where the costs of divorce are particularly high.

On the other hand, there are reasons to expect that marital discord may be associated with the likelihood of dissolution in such a setting. A 42-year-old man identified this connection between discord and marital dissolution, saying “If [a husband and wife] can’t unite their decision, if there is no understanding between them then there comes a conflict. And the conflict may [result in] a separation or divorce.” Discord can decrease spouses’ marital satisfaction—a key mechanism through which discord may lead to marital dissolution. As marital satisfaction declines, the husband or wife may consider alternatives to remaining married. If an alternative is perceived to be more beneficial than remaining married, they may end the marriage.

Spouses’ Perceptions of Discord

Spouses’ perceptions of how often they endure discord are likely to have important influences on their marital trajectories. Past research has provided evidence that wives and husbands hold different perceptions of shared experiences, such as how their marriage was formed (Bernard, 1982), the amount that each spouse contributes to the housework (Hochschild & Machung, 1989; Wilkie, Ferree, & Ratcliff, 1998), the intensity of discord (Amato and Rogers, 1997; Gottman, 1994; Matthews et al., 1996), and—in the case of marital dissolution—what caused their marriage to fail (Stewart, Copeland, Chester, Malley, & Barenbaum, 1997). By extension, we might expect wives and husbands to hold different perceptions of the frequency of discord in their marriage.

Even if spouses’ perceptions of discord are similar, they may have different abilities or incentives to dissolve the marriage. Rural Nepalese women have few economic options outside of marriage (Acharya, Bell, Simkhada, Teijlingen & Regmi., 2010), so wives may be more willing to endure discord than their husbands (Sanchez & Gager, 2000 ; Strube & Barbour, 1983). By contrast, husbands tend to hold a great deal of power relative to their wives (Bennett, 1983; Chapagain, 2006). Men have liberties that women do not, such as the means to own land and to more readily remarry (Holden, 2008). Given their lower costs to marital dissolution, men likely have a lower threshold in deciding to end their marriage. For these reasons, marriages in which husbands perceive more frequent disagreements are expected to have higher rates of dissolution.

But, there are also some reasons that wives may have incentives for seeking separation from discord-ridden marriages. Due to the practice of polygamy (Deuba & Rana, 2001), a husband is able to bring a second spouse into the home as an alternative to ending a discord-ridden marriage—an alternative unavailable to women in this part of Nepal. If a wife acquires an alternative partner while married then her motivation to dissolve her marriage is increased. Unlike a husband in the same situation, she is expected to first dissolve her current marriage before remarrying. Thus, there are reasons to expect that wives’ perceptions of discord will influence spouses’ odds of dissolution.

METHOD

Detailed, couple-level measures of marital discord and individuals’ marital experiences are unusual in South Asian settings. This study uses data from the CVFS, conducted in rural Nepal, which combine such measures with extensive panel data on marital dissolution. The data collection began in 1996 with the fielding of 72-minute, face-to-face baseline interviews. These interviews were conducted with all household members, aged 15–59 and their spouses (even if outside this age range or living elsewhere), of every household in 151 sampled neighborhoods. Special care was taken to interview spouses simultaneously in two separate locations to enhance the independence of their responses.

Following the 1996 baseline interview, original respondents began completing monthly interviews as part of a household registry that was initiated in 1997. These monthly interviews collected information on family and life events such as separation and divorce. Information about marital status was collected from interviews with one member of the household who reported on the experiences of all household members. This household registry offers benefits over court-based data on registered divorces, which exclude the many marital dissolutions that are not formally reported or legally registered (Dommaraju & Jones, 2011). Reports, or perceptions, of marital discord from the 1996 baseline interview are used to predict marital dissolution with 13 years of data from the household registry.

The analytic sample includes all couples who were married at the time of the baseline interview and in the first month of the household registry (total base sample of n = 1,570) in which wives were aged 14 to 31 in 1996 (reducing the sample by 51%, to n = 777) and in their first marriage (reducing the sample by 4%, to n = 748), and whose husband was also interviewed in 1996 (reducing the sample by 7%, to n = 698). One of the husbands in the sample had two wives who met these sample restrictions. In this polygamous case, I follow that husband’s marriage to his first wife, only, and not his marriage to his second wife. Excluding missing values, the analytic sample includes a total of 674 couples. The analysis follows the couples’ monthly hazard of marital dissolution for 162 months. The sample is limited to couples in which wives were ages 31 and younger because the experience of marital dissolution was very rare and infrequent for couples in which wives were above this age range. Some couples in which wives were ages 31 or younger in 1996 may have dissolved prior to 1996, creating some left censoring. Yet, the higher rate of marital dissolution for this sample maximizes the opportunity to examine the influence of marital discord on marital dissolution. Even in restricting the sample to couples experiencing the highest rate of dissolution, however, the rate is low: 5% of the 674 couples experienced marital dissolution during the 162 months.

Measures

Dependent

The concept of marital dissolution is operationalized by combining the events of separation and divorce: a common approach (Hirschman & Teerawichitchainan, 2003; Morgan, Lye & Condran, 1988; Morgan & Rindfuss, 1985; South, 2001). Combining separation and divorce into a single event allows pinpointing of the time at which the marriage was first disrupted. This is especially important in a setting where separation can often occur without a divorce to follow (Dommaraju & Jones, 2011). On the other hand, separation is not a prerequisite for divorce in this setting, and many dissolutions are the result of immediate divorce. Of those couples in the analytic sample who experienced marital dissolution, only 31% (n = 11) initially experienced separation compared to the 69% (n = 24) who experienced immediate divorce. The measure of marital dissolution accounts for marital breakdown, exclusively, and not separation that is the result of temporary migration.

Following previous research on divorce in Asia (Hirschman & Teerawichitchainan, 2003), this study focuses on dissolution of first marriages. In Nepal nearly everyone experiences first marriage (Yabiku, 2002), but remarriage is rare. Only about 7% of ever-married women ages 40 and older in the CVFS sample had been married more than once as of 2008. The percentage is greater for men (24% had been married more than once), likely due to the practice of polygamy. Later marriages tend to be less institutionalized than first marriages in Western settings (Cherlin, 1978; Holden, 2008) and, given their rarity, are likely to be even less institutionalized in this setting (Parry, 2001). Additionally, literature on Western contexts has demonstrated that remarriages tend to have substantially different causes and are prone to a greater likelihood of dissolution than first marriages (Becker et al., 1977; Bramlett & Mosher, 2002; Cherlin, 1978), indicating that remarriage may be more selective on individual characteristics than first marriages. Restricting the focus to first marriages avoids potential biases that later marriages might introduce.

This dependent measure uses 162 months of data from the household registry to operationalize the monthly hazard of marital dissolution in discrete time. The measure of marital dissolution is coded as 0 for every month the couple is married and 1 for the first month in which the couple becomes separated or divorced, after which the couple ceases to contribute to couple-months of exposure to risk of marital dissolution. Widowhood is treated as a competing risk, so that couples in which a spouse dies cease to contribute couple-months to the hazard.

Independent

Marital discord is operationalized with a measure of spouses’ perceptions of disagreements. This measure comes from an item in the baseline interview that asks spouses how frequently they disagree with one another: “How often do you have disagreements with your (husband/wife)? Frequently, sometimes, seldom, or never?” This measure is coded from 1 to 4, with 1 indicating never and 4 indicating frequently.

Although this measure is based on spouses’ individual reports of shared experiences, the correlation between husbands’ and wives’ perceptions of disagreements is not high, at r = 0.24. This low correlation implies that, in fact, husbands and wives have unique perceptions of marital discord within the same marriage.

Controls

Many other factors have the potential to influence both discord and dissolution. These factors include nonfamily experiences, such as education and work experiences (Heuveline & Poch, 2006; Kalmijn, Graaf & Poortman, 2004; Oppenheimer, 1994; Teachman, 2002); marital experiences, such as age at marriage, marital duration, marital cohabitation, and—in this setting—participation in spouse choice (Becker et al. 1977; Hirschman & Teerawichitchainan, 2003; Morgan & Rindfuss, 1985; Niraula & Morgan, 1996; South, 2001; Teachman, 2011); and marital fertility (Bose & South, 2003; Morgan, et al., 1988; Waite & Lillard, 1991). I use both time-varying and time-invariant measures to investigate these factors. All time-varying independent measures are precise to the month and lagged by one month.

Nonfamily experiences of both the husband and wife, within each couple, are included in the models. Husband’s and wife’s educational attainment are coded as the number of years of education that each spouse completed as of 1996. Because few people in the sample received levels of education beyond eleven years, these measures are top-coded at 11 years of education. Husband’s work experience is included as a measure of salaried labor, coded as a dummy variable to indicate whether the husband ever held a salaried job as of 1996. Women rarely hold salaried jobs, and so wife’s work experience is coded as a dummy variable to indicate whether the wife ever worked for pay as of 1996.

The models also control for marital experiences that may influence a couples’ odds of dissolution. Husband’s age at first marriage and wife’s age at first marriage (correlated at r = 0.29, p < .001) are included as continuous measures. Additionally, the models include measures to indicate whether each spouse participated in selecting their husband or wife. Individuals who exercise greater spouse choice may receive less marital support from their relatives and may place greater value on personal fulfillment and satisfaction. This variable is coded on a scale from 1 to 5, from having no choice of their (first) spouse (1) to having complete choice (5) (see Ghimire et al., 2006 for information on this measure). Because spouses within the same marriage may have experienced different participation in spouse choice, the models account for separate measures, reported by both the wife and the husband.

I use a spline function to measure time-varying marital duration, which creates two distinct groups—couples married for five or fewer years, and couples married for more than five years (precise to the month)—and allows for different slopes to better reflect the rate of marital dissolution for couples in these two groups (see Courtney, McMurtry, & Zinn, 2004 for a description of piecewise linear spline hazard models). I place the knot at five years because models testing different placements revealed this five year placement to have the strongest association with marital dissolution.

Next, the models include a measure of marital cohabitation. It has become common for men in Chitwan to leave their families temporarily to earn money in a separate location (Williams, Ghimire, Axinn, Jennings, & Pradhan, 2012), and this time spent apart may cause marriages to deteriorate. Marital cohabitation is a time-varying measure, coded 1 in the months that the couple lives together and 0 in the months they do not live together. Previous marital experience can also play an important role in marital outcomes (Becker et al., 1977; Bramlett & Mosher, 2002). The models include a dummy measure indicating whether the husband had been married more than once as of 1996 (couples in which the wife was previously married are excluded from the analytic sample). A couple’s fertility experience is operationalized as their number of children, coded as a time-varying covariate to indicate the total number of children that the couple had in each month. Sensitivity analyses were also performed that excluded couples in which the husband had been married more than once, and, in separate analyses, that controlled for number of sons instead of number of children. Both sets of sensitivity analyses revealed similar results to those presented below.

The structure of family arrangements in this setting points to the importance of considering some other, setting-specific factors that have the potential to influence marital discord and marital dissolution. The models account for characteristics of the marital home that include a series of measures taken from the 1996 baseline interview. Farmland ownership is coded as a dummy variable, with a value of 1 indicating the household owns farmland, and 0 indicating that the household does not own farmland (as of 1996). Household farmland ownership is an indicator of wealth, and a woman may be motivated to stay in a household that has greater wealth. Furthermore, women in landless households tend to work outside the home for compensation (Cameron, 1998), potentially eliminating some of their perceived economic disincentive to divorce.

Place of marital residence is important to take into account in such a patrilocal setting. Distance of the marital home from the wife’s natal home is a time-invariant covariate, taken from the 1996 baseline interview, coded on a scale of 1 to 4. A value of 1 indicates that the couple lives with the wife’s parents, 2 indicates that she lives in the same village as her parents, 3 indicates that she can reach her parents’ house in one day, and 4 indicates that it takes her longer than one day to reach her parents’ house. Women who move a greater distance from their own natal home upon marriage may have less access to the support of their family and friends in seeking dissolution (Hirschman & Teerawichitchainan, 2003). Likewise, couples who live with the husband’s parents may have greater support to either maintain their marriage and improve its quality, or to seek dissolution. A measure of whether the couple lives with the husband’s parents is coded as 1 if the husband reported living with his parents in 1996 and 0 if he reported not living with his parents.

I also include measures of ethnicity and birth cohort—key demographic characteristics in this setting. Ethnicity, which is associated with both caste and religion in Nepal, is extremely important in all aspects of social life (for detailed descriptions of the different ethnic groups, see Bennett, 1983; Cameron, 1998; Fricke, 1986; & Guneratne, 2002). Upper Caste Hindus tend to be most strict in following Hindu customs (Bennett, 1983; Stash & Hannum, 2001). Thus, couples of these high caste groups may endure especially intense pressure for their marriages to succeed. Other ethnic groups have less strict marital customs to which to adhere (Cameron, 1998; Fricke, 1986; Niraula & Morgan, 1996) and, thus, may face fewer obstacles to dissolving their marriages. Models include four dummy variables to indicate wives’ ethnicity: Dalit (or lower caste Hindus), Hill Indigenous, and Terai Indigenous, with Brahmin/Chettri (or upper caste Hindus) as the reference category. Most marriage are intra-ethnic: only 13 (2%) of the wives in the analytic sample have a different ethnic identity than their husbands.

Spouses’ birth cohort is important because younger individuals have had more experience with the recent social changes and have greater exposure to Western perspectives about marriage and divorce (Axinn and Yabiku, 2001; Barber & Axinn, 2004). Younger cohorts, therefore, may be more likely to consider divorce in the case of an unhappy marriage. Husband’s and wife’s birth cohorts are coded into dummy measures, coded 1 if born before the year 1983 and 0 if born in or after 1983.

Analysis

I use discrete-time multilevel event history analysis and logistic regression to model the monthly risk of marital dissolution. The models use couple-months of exposure as the unit of analysis, with standard errors adjusted for clustering within neighborhoods to account for the clustered sampling design of the CVFS. Couples who are exposed to the risk of marital dissolution are defined as those in which wives were in their first marriage during the baseline interview (in 1996), and they do not return to the risk set after a first dissolution. As shown in Table 1, the couples included in this sample had been married for an average of about 8 years (92.80 months) at the beginning of the observation period.

Table 1.

Descriptive Statistics

Measure Mean Standard
Deviation		 Range
Marital dissolution (proportion) .05
Perceptions of Disagreements
  Husbands’ perception of frequency 1.80 0.73 1 – 4
  Wives’ perception of frequency 1.80 0.70 1 – 4
Nonfamily Experiences
  Husband’s educational attainment 6.22 4.04 0 – 11
  Wife’s educational attainment 3.81 4.00 0 – 11
  Husband ever had a salaried joba .68 .47 0 – 1
  Wife ever worked for wagesb .43 .50 0 – 1
Marital Experiences
  Husband’s age at marriage 21.08 3.79 9 – 45
  Wife’s age at marriage 16.87 2.46 9 – 28
  Husband’s level of spouse choice (had more choice) 2.99 1.78 1 – 5
  Wife’s level of spouse choice (had more choice) 2.11 1.69 1 – 5
  Marital duration, in months (first month of hazard) 92.80 55.42 1 – 253
  Marital cohabitation (first month of hazard) .71 .46 0 – 1
  Husband married more than oncec .13 .34 0 – 1
Fertility Experiences
  Number of children born (first month of hazard) 2.08 1.43 0 – 7
Characteristics of Marital Home
  Household owns farmlandd .81 .40 0 – 1
  Distance from wife’s natal home (greater distance) 2.82 0.58 1 – 4
  Couple lives with husband’s parentse .64 .48 0 – 1
Demographicse
Ethnicity
  Brahmin/Chettri (upper caste) .48 .50 0 – 1
  Dalit (low caste) .12 .32 0 – 1
  Hill Indigenous .16 .37 0 – 1
Terai Indigenous .24 .43 0 – 1
Husband’s cohort
  Cohort born before 1983 .70 .46 0 – 1
  Cohort born 1983 or later .30 .46 0 – 1
Wife’s cohort
  Cohort born before 1983 .38 .49 0 – 1
  Cohort born 1983 or later .62 .49 0 – 1
Total couples in sample 674
Total experiencing marital dissolution 35
Open in a new tab
a

Husband ever had a salaried job: 1 = ever, 0 = never.

b

Wife ever worked for wages: 1 = ever, 0 = never.

c

Husband married more than once: 1 = yes, 0 = no.

d

Household owns farmland: 1 = yes, 0 = no.

e

Couple lives with husband’s parents: 1 = yes, 0 = no.

e

For all demographic measures, a code of 1 indicates membership in that group, and a code of 0 indicates non-membership.

Results are presented as both log odds and odds ratios, but I discuss them as odds ratios. These odds ratios can be interpreted as the amount by which the odds are multiplied for each unit change in the respective independent variable. If the odds ratio is greater than 1, the effect is positive, meaning that marital dissolution occurs at a higher (faster) rate; if it is less than 1, the effect is negative, meaning that marital dissolution occurs at a lower (slower) rate. Moreover, these ratios can be easily transformed into percent change in the odds associated with each unit change in the respective independent variable by subtracting 1 from the odds ratio and multiplying by 100 (Thornton, Axinn, Xie 2007, pp. 352–353). Because so few marital dissolutions occur in each yearly interval, the yearly odds of marital dissolution are comparable to the rate of marital dissolution. For this reason, I sometimes discuss the rate of marital dissolution as interchangeable with the odds of marital dissolution. Based on unidirectional theories regarding the expected influence of marital discord, these measures are tested with one-tailed tests of significance. Control measures are tested with two-tailed tests of significance. Because each couple can have at most one event, the standard errors are accurately estimated (Allison 1982).

RESULTS

Table 2 displays the results of the event history analyses. Model 1 reveals the influence of husbands’ perceptions, net of the effects of the control measures. This model offers an initial test of two separate, but related, expectations: First, that marital discord is associated with marital dissolution in this setting and, second, that couples in which husbands perceive more frequent disagreements will dissolve at higher rates than couples in which husbands perceive less frequent disagreements. Although the overall rate of marital dissolution is low among this sample, the influence of this marital discord measure is positive, large and significant, offering confirmation that marital discord is associated with marital dissolution in this setting. Furthermore, the model reveals evidence that couples in which husbands perceive more frequent disagreements dissolve more quickly than couples in which husbands perceive less frequent disagreements. An odds ratio of 1.63 indicates that the rate of marital dissolution increases by 63% for each unit increase in husbands’ perception. With the measure of husbands’ perceptions ranging from 1 to 4, this influence is substantial.

Table 2.

Hazard Estimates of Marital Dissolution: Odds Ratios from Logistic Regression of Spouses’ Perceptions of Disagreements

Model 1 Model 2 Model 3
B SE B OR B SE B OR B SE B OR
Perceptions of Disagreementsa
  Husbands’ perception of frequency 0.49* 0.26 1.63 0.38* 0.20 1.46
  Wives’ perception of frequency 0.45** 0.18 1.56 0.40* 0.19 1.49
Controls
Nonfamily Experiences
  Husband’s educational attainment −0.09 0.08 0.91 −0.10 0.06 0.90 −0.09 0.06 0.92
  Wife’s educational attainment −0.10 0.08 0.91 −0.10 0.06 0.90 −0.10 0.06 0.90
  Husband ever had a salaried job −0.07 0.49 0.93 −0.18 0.34 0.83 −0.17 0.36 0.85
  Wife ever worked for wages −0.14 0.48 0.87 −0.00 0.36 1.00 −0.10 0.37 0.91
Marital Experiences
  Husband’s age at marriage 0.09 0.07 1.09 0.12* 0.05 1.12 0.11* 0.05 1.12
  Wife’s age at marriage −0.09 0.10 0.91 −0.02 0.07 0.98 −0.05 0.07 0.95
  Husband’s level of spouse choice (had more choice) 0.16 0.13 1.17 0.16 0.10 1.17 0.17 0.10 1.19
  Wife’s level of spouse choice (had more choice) −0.01 0.13 0.99 0.02 0.09 1.02 0.02 0.10 1.02
  Marital duration: 5 years or less −0.03 0.02 0.97 −0.02 0.01 0.98 −0.03* 0.01 0.97
  Marital duration: More than 5 years −0.00 0.01 1.00 −0.00 0.00 1.00 −0.00 0.00 1.00
  Marital cohabitation −1.96*** 0.50 0.14 −1.94*** 0.32 0.14 −1.94*** 0.34 0.14
  Husband married more than once 1.54** 0.58 4.68 2.05*** 0.43 7.77 1.93*** 0.44 6.92
Fertility Experiences
  Number of children born −0.52* 0.23 0.60 −0.54*** 0.16 0.58 −0.53** 0.16 0.59
Characteristics of Marital Home
  Household owns farmland −0.66 0.48 0.52 −0.75 0.39 0.47 −0.77* 0.39 0.46
  Distance from wife’s natal home (greater distance) −0.20 0.27 0.82 −0.22 0.19 0.80 −0.20 0.20 0.82
  Living with husband’s parents −0.49 0.49 0.61 −0.19 0.36 0.82 −0.33 0.37 0.72
Demographics
Ethnicity: Brahmin/Chettri (reference)
  Dalit 0.51 0.62 1.67 0.04 0.53 1.04 0.20 0.53 1.22
  Hill Indigenous 0.32 0.57 1.37 0.11 0.43 1.11 0.21 0.44 1.23
Terai Indigenous −1.96 1.01 0.14 −2.28** 0.74 0.10 −2.07** 0.75 0.13
Husband’s cohort: Cohort born between 1965 and 1972 (reference)
  Cohort born between 1973 and 1982 0.14 0.67 1.15 0.84 0.48 2.31 0.57 0.50 1.77
Wife’s cohort: Cohort born between 1965 and 1972 (reference)
Cohort born between 1973 and 1982 −0.36 0.71 0.70 0.21 0.49 1.23 0.01 0.50 1.01
N (couple-months) 98,001 98,001 98,001
n (couples experiencing marital dissolution) 35 35 35
Open in a new tab

Note: One-tailed tests used for disagreement measures, two-tailed tests used for controls. OR=Odds Ratio.

a

Perceptions of disagreements are scored from 1 for never to 4 for frequently.

*

p<.05.

**

p<.01.

***

p<.001.

Model 2 continues to test the association between marital discord and dissolution, analyzing whether wives’ perceptions of discord influence marital dissolution. The model reveals that wives’ perceptions of disagreements have a strong, positive influence on couples’ hazard of marital dissolution. The magnitude of this influence resembles the magnitude of that for husbands (Model 2): The rate of a couple’s dissolution increases by 56% with each unit increase in wives’ perceptions of disagreements. Like husbands’ perceptions, wives’ perceptions of discord have a large and significant influence on couples’ odds of dissolving.

Model 3 tests whether husbands’ and wives’ perceptions of more frequent disagreements have independent influences on their rate of dissolution. The model reveals that both wives’ and husbands’ perceptions maintain strong influences, even while accounting for the other spouse’s perceptions. Net of their husbands’ perceptions, wives’ perceptions of disagreements increase the rate of couples’ marital dissolution by 49% with each unit increase in their perceived frequency, and husbands’ perceptions increase the rate by 46%, net of their wives’ perceptions.

Because the distribution of perceived frequency of disagreements is skewed, models were also tested (not shown) using a dichotomous measure of disagreements (never versus ever disagree). Those results revealed similar influences, though one-unit increase in this collapsed measure is, of course, associated with a greater rate of marital dissolution compared to the 4-point scale.

Additional sensitivity analyses were performed to test the robustness of these findings. As one test, analyses were performed on the 674 couple observations, with marital dissolution treated as time-invariant. In these models, the value of each of the independent and control measures in 1996 were used to predict marital dissolution by the end of the 13-year observation period. These models reveal similar results to those obtained using hazard models and couple-months of observation. Namely, wives’ and husbands’ perceptions of more frequent disagreements are associated with a greater likelihood of marital dissolution before the end of the 13-year period. As another sensitivity test, analyses were performed with the inclusion of a measure to indicate how happy the wife reports her relationship with her mother-in-law to be (reported in 1996). Because of missing data on this measure (e.g., due to wives with deceased mothers-in-law), the sample size drops to 606, with 31 marital dissolution events. These analyses revealed no significant influence of wives’ relationship with their mothers-in-law, and accounting for this measure does not change the results shown in Table 2. Finally, models predicting divorce and separation, separately, were also tested. Although the number of events is reduced, thus reducing the predictive power, results revealed that discord predicts divorce and separation similarly.

Some of the control measures are relevant in predicting marital dissolution, as well. Marital experiences are also associated with the rate of dissolution in the expected direction. In Models 2 and 3, husbands’ age at marriage and participation in spouse choice are positively associated with marital dissolution. The rate of marital dissolution is greatly reduced when spouses are living together (marital cohabitation), but the rate increases for couples in which husbands have been married more than once. Fertility also has an important impact: Couples with more children experience a lower rate of dissolution. Finally, relative to Brahmin and Chettris, Terai Indigenous couples are less likely to experience marital dissolution (in Models 2 and 3)—an unexpected result that may indicate changing norms of marriage among Brahmin/Chettri groups.

CONCLUSION

This paper has investigated the influence of marital discord on marital dissolution in a rural, agrarian setting of South Asia, where marital dissolution is still an uncommon phenomenon, but likely to be on the rise. The results reveal that marital discord increases couples’ rate of marital dissolution, independent of the influences of other relevant experiences and individual characteristics. Although marital dissolution is relatively uncommon and stigmatized in this setting, the influence of marital discord is similar to Western settings, where divorce has been common for decades (Amato & Rogers, 1997; Cherlin, 2009; DeMaris, 2000; Gottman, 1994; Matthews et al., 1996). Moreover, both husbands’ and wives’ perceptions of discord have important influences on the odds of couples’ dissolution.

Although the finding that spouses’ perceptions have separate influences on a variety of outcomes, including divorce, is not new (Allendorf, 2007; Amato and Rogers, 1997; Glass and Fujimoto, 1994; Gottman, 1994; Wilkie et al., 1998), this finding is particularly significant in this context. It is noteworthy that marriages are more likely to end if wives perceive more frequent discord, even though women in Nepal generally have relatively little power in households and little means to support themselves economically. Moreover, the present findings reveal that wives’ perceptions of discord have an important influence on marital dissolution that is independent of their own husbands’ perceptions.

Due to the limited liberties available to Nepalese women in this setting—as compared to men—wives’ perceptions were not anticipated to maintain a strong influence on marital dissolution when accounting for their husbands’ perceptions. Wives in discord-ridden marriages are in particularly precarious situations because they can face financial hardship in dissolving their marriage (Cain, 1986; Smock, Manning, & Gupta, 1999), but remaining married can have its own negative consequences for their emotional and, sometimes, physical well-being (Naved, Azim, Bhuiya, & Persson, 2006; Finchman, Beach, Gordon & Osborne, 1997). Yet, the present findings suggest that wives who perceive more frequent disagreements are somehow getting out of those marriages.

Because employment options for women are limited, their alternatives to remaining in a discord-ridden marriage are likely to involve securing support from another person to ensure their livelihood after dissolution. One possibility is that a wife may find another romantic partner, while still married to her first husband, and arrange to marry the new partner upon dissolving her first marriage. Another possibility is that a wife may plan to move in with her parents and regain support from them after dissolving her marriage. A third possibility is that a wife may opt for separation, rather than divorce, and maintain financial support from her husbands while residing separately.

In combination with these options for wives as alternatives to remaining in discord-ridden marriages, wives also face low barriers to legal divorce, relative to husbands. Divorce laws now offer protections for women (Manzione 2001) and there is a sense that laws make it more difficult for husbands to file for divorce than for wives to file. During an interview, one 32-year-old man talked about how “[Nepali] law has given priority to women. It’s a one-sided law, the rights [are] given to the women. If a woman wants then she can get divorce [from] her husband, [even if] the husband loves her it doesn’t matter.” Combined with the potential for wives to find the alternatives to remaining in discord-ridden marriages, these lower legal barriers may be reducing the costs of dissolution and allowing wives who perceive frequent discord in their marriages to leave those marriages. In fact, records from the District Court of Chitwan indicate that wives file the vast majority of divorces: Of the 529 divorces that were filed in the Chitwan District Court between 1990 and 2004, 96% of were filed by wives.

This paper has revealed an important association between spouses’ perceptions of marital discord and marital dissolution in rural Nepal. Nevertheless, there are also limitations to this investigation. First, the data do not reveal which spouse initiated the observed marital dissolutions. For example, it may be the case that a wife perceives frequent discord, but her husband may ultimately choose to end the marriage (or vice versa). Although records from the District Court of Chitwan suggest that wives file the majority of requests for divorce, it would be useful to have information on every dissolution experienced by the couples analyzed. Second, the measures of marital discord are from a single time point, and do not capture the changing dynamics of spousal disagreements. Repeated measures of perceived disagreements would likely reveal stronger influences on marital dissolution. Nonetheless, the single time point measure of discord has the power to predict marital dissolution over the subsequent 13 years. Third, modeling the influence of discord with a measure of disagreements is not comprehensive, as other types of discord are also likely to be relevant (Amato and Rogers, 1997; DeMaris, 2000; Gottman, 1994; Porter & O’Leary, 1980). In fact, access to more comprehensive measures might provide greater insight into couple dynamics and would likely reveal even stronger effects of discord on dissolution (Gottman, 1994). Fourth, these models do not account for discord with other family members. Even though sensitivity tests that included wives’ reported relationship with mothers-in-law did not change the results, it is possible that spouses’ perceptions of disagreements in the marriage reflect unmeasured discord with other household members. Fifth, the small number of marital dissolution events that occurred during the period of observation limits the power with which the statistical investigation can be performed. However, the significant results that we find with this small number of events suggest an important association between discord and dissolution.

The present findings have significant implications for international family research. Marital discord may have an influence on marital dissolution across settings, and even in places where women have seemingly bleak options after a marital dissolution. Yet, marital quality is often overlooked as an important factor in the well-being of families in poor, agrarian settings. With marital dissolution having significant consequences for women, men, and children in this kind of setting, it is crucial that researchers and policymakers focus on preserving marital quality.

Acknowledgments

I am grateful for support from the Population Studies Center at University of Michigan (grant numbers R24 HD041028 and T32 HD007339), from the Carolina Population Center at the University of North Carolina (grant numbers T32 HD007168 and R24 HD050924), and from the National Science Foundation (grant number OISE 0729709). I would like to thank the Institute for Social and Environmental Research in Chitwan, Nepal for collecting the data used here; William Axinn, Jennifer Barber, Dirgha Ghimire, Abigail Stewart, Keera Allendorf, Sarah Hayford, Rachael Pierotti, Katherine Lin, Sara Crider, and anonymous reviewers for helpful comments on earlier versions of this paper; and Cathy Sun for assisting with data management. All errors and omissions remain the responsibility of the author.

REFERENCES

  1. Acharya DR, Bell JS, Simkhada P, Teijlingen ER, Regmi P. Women’s autonomy in household decision-making: a demographic study in Nepal. Reproductive Health. 2010;7:15–27. doi: 10.1186/1742-4755-7-15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Allendorf K. Do women’s land rights promote empowerment and child health in Nepal? World Development. 2007;35:1975–1988. doi: 10.1016/j.worlddev.2006.12.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Allison PD. Discrete-time methods for the analysis of event histories. In: Leinhardt S, editor. Sociological methodology. San Francisco: Jossey Bass; 1982. [Google Scholar]
  4. Amato PR, Rogers SJ. A longitudinal study of marital problems and subsequent divorce. Journal of Marriage and the Family. 1997;59:612–624. [Google Scholar]
  5. Axinn WG, Yabiku ST. Social change, the social organization of families, and fertility limitation. American Journal of Sociology. 2001;106:1219–1261. [Google Scholar]
  6. Barber JS, Axinn WG. New ideas and fertility limitation: The role of mass media. Journal of Marriage and Family. 2004;66:1180–1200. [Google Scholar]
  7. Becker GS, Landes EM, Michael RT. An economic analysis of marital instability. Journal of Political Economy. 1977;85:1141–1187. [Google Scholar]
  8. Bennett L. Dangerous wives and sacred sisters: Social and symbolic roles of high-caste women in Nepal. New York: Columbia University Press; 1983. [Google Scholar]
  9. Bernard J. The future of marriage. New Haven, CT: Yale University Press; 1982. [Google Scholar]
  10. Bose S, South SJ. Sex composition of children and marital disruption in India. Journal of Marriage and Family. 2003;65:996–1006. [Google Scholar]
  11. Bramlett MD, Mosher WD. Vital Health Statistics, Series 23. 22. Hyattsville, MD: National Center for Health Statistics; 2002. Cohabitation, marriage, divorce, and remarriage in the United States. [PubMed] [Google Scholar]
  12. Cain MT. The consequences of reproductive failure: Dependence, mobility, and mortality among the elderly of rural South Asia. Population Studies. 1986;40:375–388. [Google Scholar]
  13. Cameron MM. On the edge of auspicious: Gender and caste in Nepal. Chicago: University of Chicago Press; 1998. [Google Scholar]
  14. Chapagain M. Conjugal power relations and couples’ participation in reproductive health decision-making: Exploring the links in Nepal. Gender Technology and Development. 2006;10:159–189. [Google Scholar]
  15. Cherlin A. Remarriage as an incomplete institution. American Journal of Sociology. 1978;84:634–650. [Google Scholar]
  16. Cherlin A. The marriage-go-round: The state of marriage and the family in America today. New York: Knopf; 2009. [Google Scholar]
  17. Cherlin A. Demographic trends in the United States: A review of research in the 2000s. Journal of Marriage and Family. 2010;72:403–419. doi: 10.1111/j.1741-3737.2010.00710.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Courtney ME, McMurtry SL, Zinn A. Housing problems experienced by recipients of child welfare services. Child Welfare. 2004;83:393–422. [PubMed] [Google Scholar]
  19. DeMaris A. Till discord do us part: The role of physical and verbal conflict in union disruption. Journal of Marriage and the Family. 2000;62:683–692. [Google Scholar]
  20. Deuba AR, Rana PS. Psycho-social impacts of violence against women and girls with special focus on rape, incest, and polygamy. Kathmandu, Nepal: Saathi and SNV; 2001. [Google Scholar]
  21. Dion KK, Dion KL. Individualistic and collectivistic perspectives on gender and the cultural context of love and intimacy. Journal of Social Issues. 1993;49:53–69. [Google Scholar]
  22. Dommaraju P, Jones G. Divorce trends in Asia. Asian Journal of Social Science. 2011;39:725–750. [Google Scholar]
  23. Finchman FD, Beach SRH, Harold GT, Osborne LN. Marital satisfaction and depression: Different causal relationships for men and women? Psychological Science. 1997;8:351–357. [Google Scholar]
  24. Fricke TE. Himalayan households: Tamang demography and domestic processes. Ann Arbor, MI: UMI Research Press; 1986. [Google Scholar]
  25. Ghimire DJ, Axinn WG, Yabiku SA, Thornton A. Social change, premarital non-family experience and spouse choice in an arranged marriage society. American Journal of Sociology. 2006;111:1181–1219. [Google Scholar]
  26. Glass J, Fujimoto T. Housework, paid work, and depression among husbands and wives. Journal of Health and Social Behavior. 1994;35:179–191. [PubMed] [Google Scholar]
  27. Goode WJ. World revolution and family patterns. New York: The Free Press; 1970. [Google Scholar]
  28. Goode WJ. World changes in divorce patterns. New Haven: Yale University Press; 1993. [Google Scholar]
  29. Gottman JM. The relationship between marital processes and marital outcomes. Hillsdale, New Jersey: Lawrence Erlbaum Associates; 1994. What predicts divorce? [Google Scholar]
  30. Guneratne A. Many tongues, one people: The making of Tharu identity in Nepal. Ithaca: Cornell University Press; 2002. [Google Scholar]
  31. Heuveline P, Poch B. Do marriages forget their past? Marital stability in post-Khmer Rouge Cambodia. Demography. 2006;43:99–125. doi: 10.1353/dem.2006.0005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Hirschman C, Teerawichitchainan B. Cultural and socioeconomic influences on divorce during modernization: Southeast Asia, 1940s to 1960s. Population and Development Review. 2003;29:215–253. [Google Scholar]
  33. Hochschild A, Machung A. The second shift. New York: Avon Books; 1989. [Google Scholar]
  34. Holden L. Hindu divorce: A legal anthropology. London: Ashgate; 2008. [Google Scholar]
  35. Jayakody R, Thornton A, Axinn W. International Family Change: Ideational Perspectives. New York: Taylor & Francis Group; 2008. [Google Scholar]
  36. Kalmijn M, De Graaf PM, Poortman A. Interactions between cultural and economic determinants of divorce in the Netherlands. Journal of Marriage and Family. 2004;66:75–89. [Google Scholar]
  37. Manzione LL. Human rights in the Kingdom of Nepal: Do they exist on paper? Brooklyn Journal of International Law. 2001;27:193–244. [Google Scholar]
  38. Matthews LS, Wickrama KAS, Conger RD. Predicting marital instability from spouse and observer reports of marital interaction. Journal of Marriage and the Family. 1996;58:641–655. [Google Scholar]
  39. Morgan PS, Lye DN, Condran GA. Sons, daughters, and the risk of marital disruption. American Journal of Sociology. 1988;94:110–129. [Google Scholar]
  40. Morgan PS, Rindfuss RR. Marital Disruption: Structural and Temporal Dimensions. American Journal of Sociology. 1985;90:1055–1077. [Google Scholar]
  41. Naved RT, Azim S, Bhuiya A, Persson LA. Physical violence by husbands: Magnitude, disclosure, and help-seeking behavior of women in Bangladesh. Social Science and Medicine. 2006;62:2917–2929. doi: 10.1016/j.socscimed.2005.12.001. [DOI] [PubMed] [Google Scholar]
  42. Niraula BB, Morgan PS. Marriage formation, post-marital conflict with natal kin and autonomy of women: Evidence from two Nepali settings. Population Studies. 1996;50:35–50. [Google Scholar]
  43. Oppenheimer VK. Women’s rising employment and the future of family instability in industrial societies. Population and Development Review. 1994;20:293–342. [Google Scholar]
  44. Parry JP. Ankalu’s errant wife: Sex, marriage, and industry in contemporary Chhattisgarh. Modern Asian Studies. 2001;35:783–820. [Google Scholar]
  45. Porter B, O’Leary KD. Marital discord and childhood behavior problems. Journal of Abnormal Child Psychology. 1980;8:287–295. doi: 10.1007/BF00916376. [DOI] [PubMed] [Google Scholar]
  46. Sanchez L, Gager CT. Hard living, perceived entitlement to a great marriage, and marital dissolution. Journal of Marriage and Family. 2000;62:708–722. [Google Scholar]
  47. Sastry J, Ross CE. Asian ethnicity and the sense of personal control. Social Psychology Quarterly. 1998;61:101–120. [Google Scholar]
  48. Smock PJ, Manning WD, Gupta S. The effect of marriage and divorce on women’s economic well-being. American Sociological Review. 1999;64:794–812. [Google Scholar]
  49. South SJ. The geographic context of divorce: Do neighborhoods matter? Journal of Marriage and Family. 2001;63:755–766. [Google Scholar]
  50. Stash S, Hannum E. Who goes to school? Educational stratification by gender, caste, and ethnicity in Nepal. Comparative Education Review. 2001;45:354–378. [Google Scholar]
  51. Stewart AJ, Copeland AP, Chester NL, Malley JE, Barenbaum NB. Separating together: How divorce transforms families. New York: Guilford Press; 1997. [Google Scholar]
  52. Strube MJ, Barbour L. The decision to leave and abusive relationship: Economic dependence and psychological commitment. Journal of Marriage and the Family. 1983;45:785–793. [Google Scholar]
  53. Teachman JD. Stability across cohorts in divorce risk factors. Demography. 2002;39:331–351. doi: 10.1353/dem.2002.0019. [DOI] [PubMed] [Google Scholar]
  54. Teachman JD. Modeling repeatable events using discrete-time Data: Predicting marital dissolution. Journal of Marriage and Family. 2011;73:525–540. [Google Scholar]
  55. Thornton A. The developmental paradigm, reading history sideways, and family change. Demography. 2001;38:449–465. doi: 10.1353/dem.2001.0039. [DOI] [PubMed] [Google Scholar]
  56. Thornton A, Axinn WG, Xie Y. Marriage and cohabitation. Chicago: University of Chicago Press; 2007. [Google Scholar]
  57. Waite LJ, Lillard LA. Children and marital disruption. American Journal of Sociology. 1991;96:930–953. [Google Scholar]
  58. Wilkie J, Ferree MM, Ratcliff KS. Gender and fairness: Marital satisfaction in two-earner couples. Journal of Marriage and the Family. 1998;60:577–594. [Google Scholar]
  59. Williams NE, Ghimire DJ, Axinn WG, Jennings EA, Pradhan MS. A micro-level event-centered approach to investigating armed conflict and population responses. Demography. 2012;49:1521–1546. doi: 10.1007/s13524-012-0134-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Yabiku ST. Unpublished Ph.D. dissertation. Department of Sociology. Ann Arbor: The University of Michigan; 2002. Social organization and marriage timing in Nepal. [Google Scholar]
  61. Yabiku ST. The effect of non-family experiences on age at marriage in a setting of rapid social change. Population Studies. 2005;59:339–354. doi: 10.1080/00324720500223393. [DOI] [PubMed] [Google Scholar]

RESOURCES