Published in the final edited form as:
Dommaraju, Premchand. 2016. " Divorce and Separation in India." Population and Development
Review forthcoming. [In Press]
Divorce and Separation in India
In India divorce and separation are perceived to be relatively rare events and available data on the
subject has been described as ―quantitative gossip‖. While quantitative data are lacking, there is rich
multi-disciplinary literature on various aspects of marital stability in India. This paper draws on this
literature to first contextualize the key aspects of marriage and the socio-cultural and legal systems
that influence durability of marriages in India. Following this, the paper presents estimates of
prevalence, trends and variations in divorce and separation using data from a large nationally
representative survey. Finally, using education as a broad measure, the paper investigates the impact
of social changes on durability of marriages between 1987 and 2007. The findings reveal an upward
trend and significant variations by region, religion, rural and urban, and number and sex of children
in divorce and separation. The findings show that marriage among higher educated are more durable
and that the difference between higher and lower educated women has widened over time.
Division of Sociology, School of Humanities and Social Sciences, Nanyang Technological University, 14 Nanyang
Drive, Singapore 637332. Email: email@example.com .
In many parts of Asia, non-marriage and divorce are becoming increasingly common. In East Asia,
with the exception of China, marriage rates have fallen and divorce rates briefly reached as high as
the levels seen in countries in the Organisation for Economic Co-operation and Development
(OECD) (Dommaraju and Jones 2011). In Southeast Asia, too, marital dissolution through divorce
or separation is rising (ibid.). In China where marriage rates have not declined to the levels seen in
other East Asian countries, divorce rates have tripled or quadrupled over the last two decades (Lu
and Wang 2014). The major exception to this trend in Asia is South Asia where marriage continues
to be nearly universal and relatively early, and divorce believed to be relatively uncommon.
In India, the focus of this paper, despite wide social changes over the last three decades,
marriage is still predominately caste endogamous, universal and early, and involves active
participation of family and kin in spouse selection. Though there have been changes in many of the
aspects of the marriage system including marriage age and spouse selection (Allendorf 2013; Andrist,
Banerji and Desai 2013; Desai and Andrist 2010; Prakash and Singh 2013), these have not radically
transformed the institution of marriage. While much is known about changes in marriage formation
in India, there is comparatively little socio-demographic research on marital dissolution in India.
With the notable exception of Bose and South‘s (2003) work on son preference and marital
dissolution, there has not been any major quantitative study on divorce in India. The presumed rarity
of divorce in India and lack of appropriate data has meant that marital dissolution has not received
This paper has three broad objectives. First, to present nationally representative estimates of
prevalence, trends, and variations in divorce and separation in India. Media reports with headlines
such as ―The great Indian wedding is succumbing to the great Indian divorce‖
, have proclaimed
dramatic increases in divorce and separation in India. These reports are, however, based on partial
and incomplete data and non-rigorous analyses. Besides media reports there are scattered
ethnographic data which by their nature are not generalizable. On the whole quantitative
information on divorce and separation are lacking. The estimates presented in this paper will provide
New York Times, 19th February 2008. Other media headlines include “In tradition-bound India, female, divorced
and happy” (Washington Post, 19th September 2008); “Not so happily ever after as Indian divorce rate doubles”
(BBC, 1st January 2011); “Divorce rates climbing up in Bangalore” (Times of India, 2nd August 2007).
firmer quantitative basis for understanding, among other things, the trends in marital stability, and
variations in marital stability by region, religion, family composition and education.
Second, the paper contributes to our understanding of how marriage stability is affected by
broader changes in the social-cultural milieu and specific changes in the marriage system. Some of
the factors that are driving divorce trends in other parts of Asia—such as greater participation of
women in the labour force, increasing incompatibility of gender roles with the status of women,
changing demographic contexts, marriage system, and individualization—are, though to a lesser
degree, occurring in India. Using educational attainment as broad measure to capture individual
autonomy, independence, economic potential, social status and wider social changes (Matysiak et al.
2014), the study investigates how these changes have affected marital stability in India.
Third, the paper presents a systematic synthesis of key elements of marriage and kinship
systems, pluralistic legal and social structure, and family and cultural norms that influence marital
stability in India. The paper uses these elements to frame, contextualize, interpret and discuss the
findings of the study. The synthesis draws on literature from multiple disciplines including legal
anthropology, social and cultural studies, and family studies.
2. Marriage and dissolution in India
2.1 Marriage: sexuality, spouse selection and marital stability
A key feature of the Indian socio-cultural and marriage system is the strict policing of sexuality and
proscription of sexuality to within marriage for women (Abraham 2001; Dutta 2011; Still 2010)
woman who indulges in pre-marital sex is considered as bringing shame to not only the family, but
also as corrupting the purity of caste. Family and caste councils, therefore, have assumed the role of
policing women‘s sexuality (Abraham 2014; Chakravarti 1993). The policing of sexuality is stricter
among groups where consciousness of caste identity is stronger (Abraham 2014; Dube 2001). The
concern about sexuality has meant that unmarried daughters are seen as a burden and a threat to the
existing marriage and social order (Kodoth 2008; Netting 2010). This concern about sexuality,
among other factors, has meant that marriage is relatively early and nearly universal.
The level of pre-marital sexual experience is reportedly low. About 4 per cent of young women age 15 to 24
reported having such an experience in a sub-national representative sample conducted in 2006-7 (IIPS and
Population Council 2010).
Second feature of the Indian marriage system that could influence marital stability is the
nature of spouse selection. Marriages are typically arranged by parents or families (with varying
degrees of consent and discussion with children) rather than based on pre-marital dating or
(Allendorf 2013; Prakash and Singh 2013). Parents and family play an important role
as marriage is seen as affecting the status of the family and lineage (Harlan and Courtright 1995).
Though arranged marriages have persisted, the degree of involvement of parents, families and the
young themselves has changed. There is now greater participation and higher degree of involvement
of young men and women in selecting their own spouses through negotiation with parents and
families. Compared to the past the wishes and desires of the couples themselves are now considered
during selection of a spouse (Netting 2010; Titzmann 2013). The use of technology in matchmaking
through matrimonial websites while offering the potential to transgress traditional boundaries seems
to have re-entrenched traditional values in spouse selection (Kaur and Dhanda 2013). Even for
those young men and women choosing their own spouse, there is considerable pressure to choose
the ―right‖ person that fits with the dominant patriarchal and parochial norms and caste identities
(Abeyasekera 2013; Ahearn 2001).
The shift towards greater choice in partner selection and ‗companionate‘ marriages in which
conjugal power relationship is less hierarchical (Fuller and Narasimhan 2008; Gilbertson 2014) could
potentially impact marriage stability. Kishwar (1999), for instance, argues that stability of arranged
marriages is due to stronger material and emotional support provided by natal family and kin. Such
support might promote reconciliation and help to heal marital differences. Marriages in which family
and kin do not play a major role might be less stable. However, this perspective is seen as
stereotyping Indian marriages. Grover‘s (2009) work, on the other hand, draws attention to the
destabilizing effect of family and kin in arranged marriages. She argues that marriages in which
families did not have a greater say in selecting spouse might be more stable. This is because in such
marriages women might not expect to receive support from the families nor access to natal homes in
times of marital conflict. Marriages in which partners have a greater choice and say might also be
stable for a different reason. Such marriages might engender stability through promotion of
emotional bonds, conjugality and intimacy (Osella and Osella 2006).
Indian Human Development Survey (IHDS) data from 2005 show that only 5 per cent of women had self-arranged
marriages with no active participation of parents or families (Andrist, Banerji and Desai 2013)
The third aspect of the marriage system that could influence marital stability is the status of
remarriages. The possibility of remarriage and acceptance of such marriages might influence marital
stability. Remarriages are not unusual; however, they are gendered: accepted for men but questioned
for women. While remarriage for women is not prohibited, the status of such marriage is lower
within Hindu customary practices. Previously married women are considered impure
remarriage for women do not hold same customary and ritual significance (Holden 2008). Though
remarriages and secondary unions (without marriage) for women do happen (Aura 2008; Grover
2011; Parry 2001), the quality of matches (in terms of age, economic situation, marital status, family
structure of the spouse) are unfavourable for most women and more so for women with
children(Mukhopadhyay 2011; White 2013). The possibilities of re-partnering and quality of matches
might weigh in on the decision of a woman to leave or stay in a marriage.
2.2 Jural and conjugal stability of marriages in India
The need to distinguish between jural and conjugal stability is especially important in the context of
India. Parry (2001) has observed that in India physical separation may not end the legal relationship
or the absence of legal separation does not guarantee that the couple will remain together. The
reasons for and extent to which the two forms of stability differ need to be placed in the legal,
customary, socio-cultural and religious contexts of India.
Jural stability of marriages
The Indian jural system on matters related to marriage and divorce is plural in two senses of the
word. First, there are different sets of laws for different groups; and, second, there is plurality of
venues for resolution of marital disputes (Vatuk 2013a). This plurality is historical rooted on the
need for consensual legislation and is in contrast with legal centrism and single legal code for family
matters in much of the west (Ghosh 2009).
There are several family laws (known as personal laws in India) based on religion and these
are legislated by the central (federal or national) government and not the state governments
This is exemplified by a saying quoted by Aura (2008: 80) “likening women to earthen water jugs which are
permeable and became easily contaminated to such an extent that they cannot be purified” and, therefore, once
married a woman becomes impure. However, this is not the case for men who are “like impermeable brass jugs
which are difficult to contaminate and easy to purify.”
Family matters are in the concurrent list of the Indian constitution which means that either the states or the central
government can legislate on family matters. However, central legislation takes precedence and supersedes any state
Different religious groups are governed by their respective codes and by customary laws of their
respective communities. Hindu marriages and divorces are governed by the Hindu Marriage Act
(henceforth HMA) which also applies to Buddhist, Jains and Sikhs
. There are separate laws
governing family matters for Muslims
. There is also a Special Marriage Act
(henceforth SMA) which allows for marriage between members of any or no religious affiliation and
governs divorces for marriages under SMA
The Hindu and Special Marriage Acts have similar grounds for granting of divorce. The
main grounds of grating divorce can be summarized under the following categories: matrimonial
fault, special circumstances and by mutual consent. The first ground of matrimonial fault includes
adultery (with stringent proof), cruelty
, desertion for not less than two years, unsound mind,
communicable venereal diseases, incurable leprosy, renunciation of the world by entering religious
orders, has not been seen alive for period of seven or more years (Garg 1998; Parliament of India
2011; Solanki 2011). For wife there are additional grounds of divorce under special circumstance
such as if the husband has been found guilty of rape, sodomy or bestiality or if the marriage was
solemnized before the woman turned 15 and she repudiates the marriage before age 18. Finally
divorce could also be granted by mutual consent (ibid.). There is no provision under HMA or SMA
to grant divorce due to irretrievable breakdown of marriage or to grant divorce without mutual
consent that does not involve fault or special circumstances.
legislation. For matters related to marriage and divorce, as there are existing central legislations, the states‟
legislative role has been limited (Garg 1998).
Hindu Marriage Act was enacted in 1955 and subsequently amended in 1976 to include cruelty, desertion and
mutual consent as grounds for divorce (Law Commission of India 1978).
Muslim family matters are governed by the Muslim Personal Law [Shari„at] Application Act of 1937 and
Dissolution of Muslim Marriages Act of 1939 (Jones 2010).
Special Marriage Act traces its root to 1872 and initially applied to only those who had no religion or faith. In 1954
the act was repealed and replaced by a new act under the same name. The act allows marriages in which both parties
are from different religions or from the same religion or no religion.This is an optional law and serves as an
alternative to different religious personal laws (Law Commission of India 2008b). The provisions for divorce for
those married under SMA are broadly similar to those in the HMA.
Christian marriages are governed by the Christian Marriage Act of 1872 and the India Divorce Act of 1869. The
provisions related to grounds for divorce largely remained unchanged for over a century. An amendment to the India
Divorce Act in 2001expanded divorce rights and made the provisions gender equal (Law Commission of India 1983;
Subramanian 2008). Previously grounds for men seeking divorce were very limited (ibid.).
Courts have differed in their interpretation of cruelty. Partying to acne problems have been ruled as form of
cruelty to the spouse and a reason for divorce (BBC 2015). The Supreme Court of India in recent judgments has
noted that cruelty must be judged on the “intensity, gravity, and stigmatic impact of it when meted out even once
and the deleterious effect of it on the mental attitude, necessary for maintaining a conducive matrimonial home”
(Law Commission of India 2009: 17). For Supreme Court‟s interpretation of cruelty and other aspects of the HMA,
see Kohli 2010.
An important aspect of HMA and SMA is that it recognizes customary divorce which can be
granted by customary ways through village councils, caste organization, quasi-legal and non-formal
institutions. HMA states that ―Nothing contained in this Act shall be deemed to affect any right
recognized by custom or conferred by any special enactment to obtain the dissolution of a Hindu
marriage…‖ with a liberal interpretation of custom
. This recognition of customary law means that
divorces can be obtained without any judicial involvement.
For Muslims in India, Sharia remains largely un-codified allowing for plural interpretations
(Ghosh 2009). The basic tenet underlying Islamic law is that marriage is considered a contract
can be terminated if the contract is valid or annulled if the contract is not valid. Repudiation of the
marriage by husband pronouncing (in oral or written form) talaq three times, for any or no cause, in
the presence or absence of wife, is binding and irrecoverable
(Choudhury 2008; Garg 1998; Vatuk
2013a). A wife can repudiate a marriage only if the husband grants her the power to do so or if the
wife has negotiated to have such powers in the marriage contract (Choudhury 2008). Wife and
husband can initiate divorce through mutual consent known as Khul (or Khula. Wife can also seek
divorce known as faskh from a qazi (religious teacher) or third parties if there is no mutual consent
(Redding 2012). Women, and only women, can also seek unilateral divorce under the Dissolution of
Muslim Marriages Act (DMMA) of 1939 under specified conditions (Jain 2005). These conditions
include apostasy, failure to provide maintenance, unknown whereabouts, cruelty, failure to perform
marital obligation without reasonable cause, impotence, insanity, severe disease or any grounds
recognized by Muslim law (Nichols 2012).
To get a divorce, parties can approach any regular court or family courts that have been set
up in many major cities. In a regular court or a family court, divorce proceedings can be brought
under any religious law or SMA and are heard by civil judges who are trained in common law and do
Custom and usage is defined as “any rule which, having been continuously and uniformly observed for a long
time, has obtained the force of law among Hindus in any local area, tribe, community, group or family: Provided
that the rule is certain and not unreasonable or opposed to public policy and provided further that in the case of a
rule applicable only to a family it has not been discontinued by the family.” [Hindu Marriage Act, 1955, Chapter 1:
The use of the word contract does not mean that Muslim marriages are only contracts. As Lemons (2013) explains,
Muslim marriages are in the border between ritual law and contractual law in Islamic jurisprudence and adjudication
of marital disputes does not rely solely on contractual law but also based on ethical, religious, social and emotional
There are debates among both legal and religious scholars on the interpretation and irrevocability of the triple
talaq (see Garg 1998; Lemons 2010). Indian courts have generally held that irrevocable divorce can only be granted
for a reasonable cause and after an attempt at reconciliation (Choudhury 2008; Subramanian 2008).
not have any special training in religious jurisprudence (Vatuk 2013b). Thus, a Hindu judge can
decide divorce under the Muslim act and vice versa. In addition to these formal venues, marriage
disputes can be settled using dispute resolution bodies supported by the government, NGOs or
religious bodies (Grover 2011; Solanki 2011). Community groups such as panchayats (village
committee or councils) can also mediate and resolve marital disputes and grant separations. Muslims,
for instance, can approach non-state dispute resolution forums such as the dar ul qaza run by All
India Muslim Personal Law Board and Imaarat Shariah in addition to state sponsored dispute
resolution forms (Lemons 2013; Redding 2012).
Family courts set up in major cities are aimed at more efficient disposal of cases than regular
courts and provide legal aid (Basu 2006; Solanki 2011). Family courts are meant to promote
reconciliation and resolution of marital disputes by developing consensus between spouses, failing
which a judge can rule on the claims and grant a divorce if ―convinced of the impossibility of
reconciliation‖ (Solanki 2011: 121)
. As observed by Vatuk (2013b), the courts see their role as
dispensing social justice rather than promoting justice for women.
The recognition of customary law means that divorce can happen without the need to go to
the judiciary (Nichols 2012). This provides a cheaper and informal way of negotiating marital
disputes and ending of marital relationships within individuals own ―sociocultural spheres‖ (Menski
2012). Though the legality of divorces obtained from non-state forums is not fully clear (Redding
2012; Vatuk 2013b); they are for all practical purposes accepted. Separations approved or granted by
non-state forums and actors or by the police and NGOs are accepted and recognized by
communities and families (Grover 2012; Redding 2012). Many of the cases in the family courts are
cases seeking to validate and reinforce customary divorces decided by non-state or informal process
at the customary level (Holden 2008; Solanki 2011). With availability of multiple avenues for
resolving marital disputes and ending a marriage, many of those seeking redress are engaged in, what
Holden (2008: 123) describes as, ―forum shopping strategies‖ and pick a strategic forum to put
pressure on the other spouse to settle unofficially and out-of-court (Berti 2011).
The Family Court Act of 1984 states that “It is the duty of the Family Court to make efforts for settlement…
[endeavor shall be made] to assist and persuade the parties in arriving at a settlement…[if] there is a reasonable
possibility of a settlement between the parties, the Family Court may adjourn the proceedings.
Conjugal stability of marriages
For the majority Hindus while marriage is held to be a sacred institution and indissoluble, divorce
and separation are accepted, recognized, justified, and a social reality (Holden 2008; Menski 2001).
Anthropological literature has documented the prevalence and extent of marital breakdown in
different regions and among different caste groups. These studies document that in some regions
and among certain caste groups divorce and separation was high, and in other regions and caste
groups (including higher castes) while divorce and separation was not high they were accepted (Aura
2008; Holden 2008; Parry 2001).
Separations can take various forms with different implications for conjugal stability. A
common form of separation which may or may not indicate a permanent break in conjugal
relationship is return of married women to her natal home. For Indian women, moving back to natal
home as a result of marital problems or disagreements or unhappiness with husband or in-laws is
considered as a ―right‖ and ―a moral and social entitlement‖ in a marriage (Grover 2009). Such a
right to the natal home, even if it is intended to be temporary, gives women bargaining power and
opportunities for reconciliation. The duration of stay at natal home and whether the husband comes
to fetch the wife and the success of subsequent negotiations might indicate breaking or healing of
marital bonds (Parry 2001; 2004). Therefore, it is not uncommon for women to move back and
forth between natal and conjugal homes, and these moves should not be seen as a step towards
dissolution of a marriage but rather as opening of possibilities for reconciliation. Move to natal
homes besides indication of marital unhappiness might also be determined by needs of labour at
natal homes (Palriwala 1991). Though women see it as a right to return to their natal home, for
many women support from the natal families will be tempered by the conditions at the natal home.
These conditions include presence of unmarried siblings whose marriage prospects might be
affected due to the return of married women to her natal home (Greenberg 2003). Also, parents
might feel that their obligation to their daughters to have ended at the time of marriage, and might
be reluctant or unable to support them.
There are, however, regional differences in the extent to which women can draw support
from their natal families. These differences are rooted in the prevailing kinship systems, culture and
socio-political structures of the regions. The differences in interrelated aspects of marriage rules,
lineage and post-marital residence explain the differing levels of support available to women from
their natal families after marriage (Chakraborty and Kim 2010). These also influence women‘s
position and bargaining strength within the marital family. In regions where patrilineal, patrilocal and
exogamous marriages are the norm such as in the north (and especially rigid in the northwest),
women‘s link to their natal family after marriage is weak. In these regions women are considered as
being transferred to husband‘s family and no longer contribute nor can expect to receive support
from their natal family (Das Gupta 2010). In the south where there is a higher prevalence of
endogamous and reciprocal cross-cousin marriages, there is greater flexibility in the relationship a
woman has with her natal family and women can expect higher degree of support from their natal
family (Andrist, Banerji and Desai 2013). In the north-east, though there is diversity within the
region, the system is less rigid and women are in a much more favourable position (Dikshit and
Dikshit 2014; Marak 2012). These differences in kinship, marriage and cultural system are also likely
to be reflected in the levels of marital instability across the regions.
Conjugal stability could be strained by childlessness and lack of sons. There is widespread
and intense desire to have children and a married woman‘s identity, status, power and self-worth are
tied to her reproductive fruitfulness (Bharadwaj 2003; Nandy 2015). Children are seen as a source of
power for women and they help to cement the bond between husband and wife. Infertility and
absence of children are considered as women‘s problems regardless of the actual cause (Mehta and
Kapadia 2008). As Riessman (2000) has observed, childlessness is seen as deviating from the
normative expectations and might break the fragile relationship between the spouses in an arranged
marriage where the couples don‘t have strong bonds at the time of marriage. The desire to have
children immediately after marriage is nearly universal and conscious fertility control seldom occurs
before the first birth (Dommaraju 2009). In this demographic context childless couples face
considerable social pressure.
It is not only the number of children but also the sex composition of children that affect
marital stability. Bose and South‘s (2003) study established that having sons inhibits divorce and that
this relationship is invariant for different groups of women. As with childlessness, not having a son
might strain marital bonds in a society with son preference. Guilmoto (2008) highlights the extent
and intensity of the manifestation of son preference in the form of skewed sex ratios in the
demographic context of declining fertility. Sex ratios rose rapidly in the 1990s before levelling and
slightly declining in the past decade (Bongaarts and Guilmoto 2015). The reasons for son preference
are varied and intertwined ranging from structural causes, marriage and kinship system, sexual
milieu, family structure, political and state system, and religious rituals and customs that require sons
(Das Gupta 2010; Jeffery 2014). Whatever the reasons, women are blamed for lack of sons, and
absence of sons in seen as a valid reason for men to re-marry (with or without separating from his
As described earlier, features of the marriage system such as spouse selection, marriage type,
and remarriages might also affect conjugal stability. A broad indicator that has an influence on many
of these aspects is women‘s education. Studies have noted that education increases women‘s choice
in selecting partners, delays entry into marriage, and creates less hierarchical power relationships
within marriage (Chakraborty 2012; Desai and Andrist 2010; Gilbertson 2014; Kalpagam 2008).
These could, in turn, (de)stabilize marriages. Indian literature on the effect of education on marital
stability is not well developed, but literature from other contexts suggests some potential linkages.
The influence of education on marital stability depends on prevailing rates of marital
breakdown and the degree of gender inequality. Studies have noted that the effect of education
shifts from positive to negative gradient as divorces become more common and marriage is de-
institutionalized (de Graaf 2006; Härkönen and Dronkers 2006; Kreager et al. 2013; Matysiak et al.
2014). A generalization that could be drawn is that in contexts where the cost of divorce, be it social,
family, economic or legal, is high the education gradient would be positive, as higher educated
women would be able or willing to bear the cost; and, as this cost comes down, the gradient would
turn negative. In some East Asia countries though the barriers and cost are high the gradient has
been negative due to contextual factors (Chen 2012; Lee 2006; Park and Raymo 2013). In these
societies with high degree of value attached to family reputation and honour, the social cost of
divorce might be higher for educated women (Chen 2012). Thus, the relationship between women‘s
education and marital stability might also depend on contextual factors such as the importance of
family reputation, class identity and meanings attached to marriage.
3. Quantifying marital stability
3.1 Registration, court records, censuses and surveys
Marriage and divorce registration were not compulsory. The personal laws for Hindus and Muslims
do not mandate registration of marriage and divorces (Aura 2008). Some state governments in
recent years have introduced legislation to make it compulsory to register a marriage for all or some
areas or groups, and, after repeated prodding by the Supreme Court, many other states since 2006
are in the process of introducing laws to make marriage registration compulsory (Law Commission
of India 2008a). However, these legislations are limited to registration of marriages and do not apply
for divorce or separations. With some exceptions
, there is no legislation or provision to register a
divorce or separation even if one wishes to do so.
Courts and dispute resolution bodies usually have records on number of cases filed, status of
the cases and information on reasons for seeking a divorce. They are not, however, required to
report this information to any government agency and these data have never been systematically
collated. However, researchers and media have compiled data from certain selected courts in some
. A common conclusion from these data is that the number of cases filed or divorces granted
has increased over the recent years. While such data are useful in showing that many men and
women do turn to courts to settle their marital disputes, they are not useful in gauging the trends in
marital dissolution. As Solanki (2011) notes, the numbers based on court records are incomplete and
often provide a misleading account of the trends. As a demographic measure, the numbers from the
court filings have a problematic numerator and an absent denominator. The number of cases filed
only covers those that have approached the court, and, even for those, the final outcome might not
have been a divorce. The absence of denominator means that increase in the population or number
of marriages or married couples is not taken into account. It is possible that increase in number of
cases filled might reflect to some extent the increase in number of married people. The court
records, nevertheless, provide valuable information on the type of people approaching legal venues
to seek resolution and the pleaded reasons for marital instability. Common reasons pleaded by
women include cruelty, physical abuse, alcohol problems and desertion
Some states including West Bengal, Bihar, Jharkhand, Orissa, Assam and Meghalaya provide for voluntary
registration of Muslim divorces under the Muslim Marriage and Divorce Registration Acts. There are no
corresponding provisions for the Hindus or other religious groups to register a divorce (Law Commission of India
Singh and Sinha‟s (2005) report from one family court in Mumbai shows an increase in number of cases filed
from 1,839 in 1991 to 2,877 by 2001. Aura‟s (2008) report based on family courts of Bangalore shows that 1,392
cases were filed in 1999. Media reports mentioned in footnote 1 also use data from one or more family courts in the
cities of Mumbai or Bangalore or New Delhi.
Solanki‟s (2011: 122-123) reading of case files reveal the range of misconduct alleged by the husband in contested
cases to include „refusal of sexual intercourse‟, „disrespectful behaviour‟, „unconcern toward children‟, „negligence
of cooking and household duties‟, „being “too modern” and “fun loving”, „frequent visits to the natal home‟.
Women‟s allegations against the husband include, „alcoholism, physical violence‟, „mental cruelty consisting of
verbal abuse‟, „failure to maintain the wife and children‟, gambling and sexual indifference. Though adultery is
often cited as a ground, the courts require “stringent proof” and it is difficult to obtain judgment on this ground.
Singh‟s (2013: chapter 3) study also note some of the same reasons as above.
The Indian censuses collect information on current marital status of all household members
as part of the household schedule. Census tables are published for each district and state and for the
entire nation; but, micro-data are not released. Census tables related to marital status give
information by sex and five-year age groups
on current marital status. All censuses, with the
exception of the 2011 census, combine divorced and separated into a single category in the tables. In
2011 census, numbers for divorce and separation are reported in two separate columns. These
numbers are useful in providing a sense of the current marital status and are discussed in a later
section. However, these numbers cannot be used to derive any of the standard demographic
measures of divorce, as they lack information on the timing of divorce or separation.
Surveys provide micro-level data on marital status. Over the last two decades, there have
been several large scale nationally representative demographic, household and health surveys in
. Many, if not all, of them have a question on marital status and age at marriage; but, with the
exception of one survey that is described in the next section, no information on duration of
marriage for marriages that the respondents indicated as dissolved by divorce, separation or
desertion. In addition none of them have detailed marital history data that includes information on
remarriages and the nature of divorce and separation. This lack of information makes it difficult to
reconstruct marital life histories. There are reports from smaller anthropological studies among
certain regions or groups that provide scattered quantitative evidence on dissolution of marriages
They are, however, limited in their relevance to national patterns and trends.
3.2 Data, analyses and sample characteristics
The primary data for analyses come from the third round of District Level Household and Facility
Survey (DLHS-3) conducted in 2007-8. DLHS is broadly similar to Demographic Health Surveys
(DHS), but with the aim of producing representative data for all districts, and with emphasis on
measuring the progress of health and other public programs (IIPS 2010). The survey covered 601
districts in 34 states and union territories and collected data from 720,320 households and 643,944
Marital status by single year age and sex is also available as an appendix table for some census years.
Notable of these include three rounds of National Family Health Surveys (NFHS: these are Indian version of
Demographic and Health Surveys), two rounds of Indian Human Development Survey (IHDS), District Level
Household and Facility Survey (DLHS) and National Sample Surveys (NSS).
Aura (2008) reviews many of these studies. In these studies, per cent of marriages that have ended in divorce or
separation ranges from 1.5 per cent in rural Karnataka to 6.5 per cent in Western Uttar Pradesh.
ever-married women in these households (ibid.). The large sample size allows for a robust
examination of a low risk event such as divorce and separation.
The data from the ever-married women‘s questionnaire on marital status forms the basis of
the analyses. The questionnaire asked women their current marital status and for those currently
widowed, separated, divorced or deserted information on how long they have been in their current
marital state. As some of the separation could be temporary, only those who are separated for more
than one year are considered as separated for the purposes of this paper. Those who report their
current marital status as divorced or deserted or were separated for more than one year are
combined into a single category. The plurality of forms of separation, discussed earlier, means that it
is both conceptually and analytically necessary to consider the various forms as a single category.
The survey also collected information on wide range of characteristics of the respondents.
For the analyses, information on educational attainment of women and husbands, year of marriage,
number and sex composition of living children, religion, region, area of residence and caste are used.
Education is classified into three categories broadly matching no or below primary level, below
secondary, and secondary and above. Marriage cohort is categorized in three categories. The number
of living children and sex composition are classified as no children, no living son and at least one
living son. This measure is time invariant as the survey did not collect detailed fertility history data.
As child bearing outside of marriage is highly unlikely in India, it is reasonable to assume that
children were born in the marriage before it was dissolved. The classification of caste follows Indian
administrative system in which thousands of castes are grouped based on either constitutional
requirements or based on mix of historical position of the groups and their socio-economic status.
Religion and area of residence are self-explanatory. Region is classified based on geographical
position of the states.
Standard survival analyses and discrete-time logistical models are used to analyse the data.
The event of interest is whether a woman experiences divorce, separation or desertion. The risk of
the event starts from the time a woman starts living with the husband
. For women who experience
the event, the duration of risk is between the year they started living with their husbands and the
Women might not start living with their husband immediately after marriage, especially if the marriage occurred
at a young age. But with increasing age at marriage, the practice of not living with husband immediately after
marriage is declining. As the risk starts from the time of living with the husband, the analyses might miss
dissolutions that could have occurred even before the couple started to live together.
year of the event. Women who did not experience the event are either censored at the year of
interview or the year of their widowhood if they are widowed. The duration of risk is measured in
years. The analysis is restricted to women who married between 1987 and 2007 (inclusive of both
years). Survival estimates are based on Kaplan-Meir method. Discrete-time model is based on the
person-year risk of event and the model is estimated using a logistic function (see Allison 2010;
Singer and Willett 2003). The duration of risk is modelled using a quadratic term
. The estimates are
presented in the form of odds ratios. All analyses are conducted using weights provided in the
There are some exclusions and caveats. Women in two outlying and sparely populated
islands of Andaman and Nicobar and Lakshadweep were not included in the analyses. The analyses
are based on women who at the time of survey were divorced, separated or deserted. This means
that those who were divorced or separated previously but had remarried by the time of the survey
would not be counted as divorced or separated. There is no information on remarriages in DLHS-3.
However, an examination of another nationally representative survey, Indian Human Development
Survey (IHDS), conducted in 2005-6, shows that 1.1 per cent of ever-married women had married
more than once, and only 0.78 per cent of currently married women were divorced or separated
. These numbers indicate very low rates of remarriages for women and suggest that they are
unlikely to have a major impact on the substantial conclusions drawn in the paper. The analysis does
not include measures for economic status. Though DLHS-3 has a wealth index, it is based on the
household status at the time of the survey. Using this measure of wealth might be problematic
especially for divorced and separated women as their current economic status is likely to be affected
by divorce or separation.
The sample characteristics for variables used in the analyses are presented in Table 1. The
table shows per cent in each category for those who are divorced or separated and those who are
not. The analysis is based on 466,041 ever-married women among whom 6,522 were divorced or
separated or deserted. Education profile for women shows that half of them had less than five years
A comparison of AIC and SC statistics for unrestricted, linear, quadratic and log form of duration suggested that
quadratic form fits best for the model.
IHDS interviewed about 40,000 households and collected information on variety of topics including marital
information from ever-married women. The survey had question on number of previous marriages, and status of first
marriage, besides questions on current marital status. However, it does not have information on duration of marriage
for marriages that ended in divorce or separation or widowhood. IHDS data is available from ICPSR:
of education in both divorced and not divorced or separated groups, with lower per cent of higher
educated among divorced or separated. The educational profile for husbands show higher per cent
of less educated among divorced or separated. The last two marriage cohorts (by marriage year) each
had roughly thirty per cent of women, and there was no major difference in per cent of women in
each of the marriage cohort for the two groups. Nearly three-fourth of those who are not divorced
or separated had at least one son compared to 45 per cent among those divorced or separated. A
third of divorced or separated women had no children compared to ten per cent in the other group.
The religious profile of those currently married broadly reflects the population profile. However
among the divorced or separated, the other religious categories made up 22 per cent. Roughly two-
thirds of women were living in rural areas in both groups. The per cent of women living in the
northeast and south were high among those who were divorced or separated. The mean marriage
age does not differ between the two groups. As would be expected, mean marital duration is shorter
(5.9 years) for those who are divorced or separated compared to those who are not (11.5 years).
[Table 1 about here]
4.1 Descriptive findings
The census of 2011 recorded nearly 1.5 and 3.2 million divorced or separated men and women
respectively (see Table 2). The number of people who were divorced or separated more than
doubled over the past two decades. The number of women divorced or separated was twice that of
men in 2011 and has been higher in all the census years since 1961. The lower number of divorced
or separated men is almost certainly due to higher rates of remarriage among men compared to
women. Per cent divorced or separated among ever married age 15 and over does not exceed one
per cent for men and women in both rural and urban areas, except for urban women where it
reached 1.04 per cent in 2011. The magnitude of the difference between rural and urban areas is not
large. The percentages show a continuous increase over the last three censuses starting from 1991.
The 2011 census presents numbers for divorced and separated separately. Among divorced and
separated, those who separated constituted about 72 per cent. This per cent was about the same for
men and women (71.9 and 72.3 per cent respectively).
[Table 2 about here]
The age patterns seen in Figure 1 reveal a picture of increasing levels of divorce or
separation by age for women peaking at the age of 40-44 in the most recent census. For men,
however, setting aside the pattern at very young ages, the age pattern does not show any marked
peaks between 25-29 and 45-49. For men aged 15-19 there is notably higher level of divorce or
separation in recent decades, reflecting the instability of early marriages for men. The levels for 2011
are higher than all other census years, except 1961, in most age groups. The reasons for higher level
seen in 1961 in almost all age groups for men need further investigation.
[Figure1 about here]
The trend in probability of divorce or separation for women based on DLHS-3 data is
presented in Figure 2
. The figure shows predicted probability of divorce or separation between
1987 and 2008. The probabilities of divorce and separation show a rising trend over the twenty year
period, with rapid increases seen after 1999-2001. The probabilities nearly tripled by the end of the
observation period. This rise roughly corresponds to social and economic changes of the early 1990s
and liberalization of various spheres of social and individual lives.
[Figure 2 about here]
The per cent of marriages that end in divorce or separation for different groups are
presented in Figure 3. These are based on survival estimates. About 2 per cent of all marriages in
India end in divorce or separation within 20 years of marriage. This varies from around four to less
than one per cent in northeast and northern regions respectively. Northeast, with roughly 4 per cent
of India‘s population, relate culturally not only to mainland South Asia but also to Southeast Asia
and East Asia (De Maaker and Joshi 2007; Dikshit and Dikshit 2014), and have distinct demographic
and marriage patterns, and gender systems. The differences seen between the south and northern
regions at the two ends mirror differences in other demographic patterns such as fertility. These
differences are partly due to differences in kinship and marriage systems and also differences in
status of women in the two regions.
Religious differences in divorce and separation show lower rates of dissolution for Hindus
followed by Muslims and the highest for the other religious group (this group includes Christians,
Buddhist and Sikhs). The biggest difference is not between Hindus and Muslims, though there is a
Probabilities are estimated using a discrete-time logit model with year and quadratic form of marital duration as
difference, but between these two groups and the other religious group. For the other religious
category, per cent of marriages ending in divorce or separation reaches four per cent by the end of
the 20 year period. Whether these religious differences are results of spatial distribution of the
groups and the prevailing contexts in the regions or due to socio-economic characteristics are tested
next using multivariate models.
Educational differences presented in Figure 3 indicate that women with higher education
have the lowest risk of divorce or separation compared to those with lower levels of education. For
those in the middle education category, the risk is even slightly higher than those with low or no
education. In the context of low rates of marital dissolution, higher educated women could be
expected to have enough resources to overcome social and other costs of divorce or separation.
However, as the estimates suggest, higher educated appear to have more stable marriages. The
reasons for this are discussed in the next section.
The final chart of Figure 3 shows survival estimates by number and sex of children. Those
with at least one son have the lowest risk of divorce or separation and the highest risk is for women
with no children. Nearly 11 per cent of those with no children have marriages ended by divorce or
separation. This difference is high compared to differences seen for all other groups in the preceding
charts. The difference between no children and those with children is striking. The risk for those
with no son compared to at least one son is higher; but, the magnitude of the difference is not as
large as between no children and at least one son. In a strongly pro-natal society where having
children is seen as an essential aspect of marriage, lack of children might severely strain marital
[Figure 3 about here]
4.2 Odds ratios
and adjusted odds ratio from the discrete-time models are presented in Table 3.
The unadjusted odds for marriage cohort indicate a pattern of increasing risk for recent cohorts.
Among other factors, urban women have higher risk than rural women and those with lowest level
of husband‘s education have higher risk. The unadjusted estimates for women‘s education, religion,
living children and region are similar to those presented in the survival estimates charts earlier.
Unadjusted models include marital duration besides the variable of interest.
The adjusted odds ratios presented in the next column confirm the direction and magnitude
of the educational differences seen earlier. Those in the middle category of education are 45 per cent
more likely than those with higher education to be divorced or separated. The difference between
the lowest and middle educational groups can also be compared (results from all comparisons are
not shown in the table). Women with no or lower education were 20 per cent less likely than the
middle education category to be divorced or separated. Women whose husbands had low or no
education are twice more likely to divorce than women whose husbands had higher education.
Lower education of women and husbands increases the risk of divorce of separation.
Absence of children increases the risk of divorce or separation ten-fold, and absence of sons
about three-fold compared to those with at least one living son. The odds of divorce or separation
for those with no children are more than three times that of those with no son (not shown in table).
Moving to marriage cohorts, the estimates in the adjusted models indicate that women who married
before 2002-2008 had lower risk of divorce and separation than those who married in 2002-2008.
Muslims and those in other religious category have higher odds (32 and 81 per cent higher)
of divorce and separation compared to Hindus; Muslims are 27 per cent less likely than the others to
divorce or separate (results not shown in table). Women belonging to Scheduled Caste and Tribe
have higher risk than other caste groups. The effect of urban region is higher in the adjusted model,
with urban women 50 per cent more likely to divorce or separate. The pattern of regional difference
seen in unadjusted models remains similar in the adjusted model. Regional comparison from the
model shows that southern states had higher risk than those in the north, east and central regions.
The only two regions that are either higher or not different from the south are the northeast and the
[Table 3 about here]
The changing role of education across marriage cohorts is examined using an interaction
term for education and marriage cohort in a model that includes all other covariates as in the
adjusted model in Table 3. The results from the interaction model are presented in Table 4. For
brevity, only the interaction estimates are presented in the table and estimates for other covariates
are not shown. The odds of divorce or separation for women with no education compared to those
with higher education changes from being not significant for the earliest marriage cohort to 40 per
cent higher in the most recent marriage cohort. Similarly, the odds for divorce or separation for
middle education group compared to those with higher education increases for each cohort, and for
the most recent cohort about 75 per cent higher. The difference between lowest and middle
educational level changes only modestly. Overall, the results suggest an increase in marital instability
for women with lower and middle education compared to those with higher education.
[Table 4 about here]
The study presented estimates of trends, prevalence, variations and factors affecting marital stability
in India. The estimates show a rising trend of divorce and separation though starting from a low
base. Many of the changes that have occurred over the last two decades might have influenced
marital stability. These changes include hybridization of the marriage system, changing role of
women, ideational shifts, access to legal system, economic growth accompanied by economic
insecurity, communication revolution, opening up of new possibilities for meeting and (re)partnering
(Baas 2009; Dommaraju and Jones 2011; Grover 2011; Jauregui and McGuinness 2003; Jesmin and
Salway 2000). These changes should not, however, be considered as necessarily a move towards
western or ―modern‖ institutions. As in other parts of Asia, while these changes could have
influenced marital stability, the direction and magnitude of the influence are context specific. One
such change in marriage system—shift from arranged marriage to hybridization involving choice—
could either stabilize a marriage due to better match or destabilize a marriage in the absence of
anchoring support of close family and kin. The results from the effect of education, used as a broad
indicator to capture some of these changes, are suggestive of the potential influence of some of
these factors in the context of India.
Women with higher education had more durable marriages than women with lower or no
education. This educational difference has widened over the two decades (1987 to 2008) with
increasing instability for women with no or low education compared to those with higher education.
Hybrid marriages and marriages in which women have a greater say are much more common among
those with higher education (Andrist, Banerji and Desai 2013). The durability of marriages among
the higher educated might be an indication that choice in marriage partner contributes to marital
The influence of education in India is in contrast with the expectation that in a low divorce
setting, where cost of divorce is high, better educated women might have resources necessary to
leave a marriage. Women with higher education might also have different expectations from a
marriage which might conflict with existing gender roles. Among the educated and middle and
upper classes, the notion of good marriage has shifted to include sharing of responsibilities, equality,
understanding, support after marriage and a greater emphasis on equality (Allendorf 2012;
Gilbertson 2014; Jaiswal 2014; Palit 2014). For this group, marriage can no longer be stereotyped as
collectivist and pragmatic. Notions of romantic love, emotional, physical and cognitive intimacy are
considered important for marital satisfaction (ibid.), and perceptions of marital discord could lead to
divorce and separation (Jennings 2014).
The durability of marriages among the higher educated points to the possibility that higher
bargaining power, greater autonomy in spouse selection, more equal gender relationship could create
strong marital bonds. This also suggests that durability of marriage might also serve as a marker for
class distinctions. For the middle and upper classes a good marriage is one that is durable, and
women are seen as responsible for maintain stable marriages (Chakraborty 2012). As Gilbertson
(2014: 238) observes, women in middle and upper class ―balances ideals of marriages characterised
by friendship, equality and emotional intimacy with ideals of the ‗good‘ Indian family to which a
durable marriage and feminine adjustment are central‖. For educated women stability of their
marriages serves as a symbol of good marriage and respectability. This stability serves as a
distinguishing marker of class identity in contrast to the unstable and permissive marriages among
the less educated and lower class. The permissiveness and acceptability of different forms and
multiple conjugal unions among the lower classes are facilitated by structural factors and
deprivations in low income neighbourhoods (Grover 2011). The intensification of structural
disadvantages and deprivations, coupled with rising aspirations in a liberalized economic structure
accompanied by labour market uncertainties and inequalities, might have further destabilized
marriages among the less educated.
The results show the importance of childbearing and having sons on marital stability. With
nearly universal marriage and primacy of children in the family ideal, lack of children in a marriage
puts tremendous strain on marital bonds. The burden of not bearing children, irrespective of actual
reason, tends to fall heavily on women. Men in a childless marriage take it as a legitimate reason to
get remarried with or without separating from the current spouse. In such a situation, break down of
marital bond is inevitable (Holden 2008; Parry 2001). Though some childless women are seeking to
construct different identities and sustain marital ties (Riessman 2000), it seems for many couples
childlessness spells the end of a marriage. The present findings affirm Bose and South‘s (2003)
conclusion a decade earlier and show the continued relevance of pro-natal and son preference ideals
on marital stability.
The results also highlight variations by region, religion and rural/urban residence. The
regional differences show higher marital instability in the northeast and south. These two regions
have different kinship and marriage systems which are less rigid and patriarchal, provide for
maintaining of ties with natal family after marriage, higher autonomy for women, and where social
and political hierarchies have been upended by socio-political transformations over the last half
century (Dommaraju and Agadjanian 2009; Xaxa 2004). These differences could partially explain the
higher rates of instability in these two regions. The findings establish the significance of religious
differentials in marital stability. These differentials could be due to differences in acceptability and
ease of divorce (at least for men), differences in position of women in the private sphere (which
could be different from the public sphere, see Desai and Gheda 2014), differences in social and
economic position of the communities, and differences in group identity and norms. These are
issues that need further examination. Higher rates seen among urban women suggest the potential
influence of rising economic uncertainties among the lower classes, greater opportunities for re-
partnering, and shifts in marital ideologies.
The comparatively low level of marital dissolution in India would be an ideal if marital
stability was underpinned by marital satisfaction and happiness, symmetrical gender power
relationship, equitable socio-legal system and egalitarian values. For many women, however, staying
or leaving a marriage is not necessarily based or dependant on any of these values. In a system where
women see leaving a marriage as their last alternative of either killing themselves or being killed by
their husbands (Aura 2008), and leaving a marriage is still seen as a ―reflection of her character,
morals, or child-bearing ability‖ (Hornbeck et al. 2007: 277), divorce or separation is a poor measure
of marital satisfaction or happiness. Besides shame, divorced or separated women face the risk of
sexual predation and harassment as a single woman (Walters et al. 2012). Divorced or separated
women have limited right for maintenance and no right to property and assets accumulated during
the marriage (Agnes 2009; Singh 2013); their right to maintenance is often symbolic and even such
symbolic amounts are not paid. Though some parents are able and willing to support their divorced
daughters (Greenberg 2003), such support is still limited. The difficulties in re-partnering also might
limit woman‘s choices. Growth in IT and online matrimonial sites for those seeking to remarry
provides new possibilities for re- partnering. Educated and urban women would be better placed to
use these opportunities. However, for the majority of women multiple forces of sexual morality,
gendered power relationship, political economy of legal and economic rights might act to prevent
women from leaving unhappy marriages.
The pluralistic nature of family law means that there are multiple venues and multiple laws
including customary ones to end an unhappy marriage. This plurality, however, does not translate to
foundational differences in views on marriage, marital stability, and women‘s role, behaviour and
responsibilities; many of the venues and laws or customs work under the same gendered and
patriarchal assumptions (Basu 2006, 2012; Vatuk 2013b). Dissolution of a marriage by approaching a
court is still limited compared to customary and informal separations. The reasons for this are not
difficult to locate—courts are seen as being ineffective and inefficient, costly (bribes, lawyer and
court fees), complex and out of reach, and due to general lack of faith in the judicial system; in
addition, approaching courts, without resolving marital disputes within the family or community is
seen as bringing shame to the family and community (Chowdhry 2004; Grover 2011; Redding 2012;
An amendment to HMA and SMA was introduced in the parliament in 2010 to make
divorce procedures less difficult by introducing irretrievable breakdown of marriage as a ground for
divorce (Law Commission of India 2009; Parliament of India 2011). This would have allowed for ex-
parte divorce without proof of fault. However, the amendment has not been passed by the
parliament due to reservations on the potential ‗misuse‘ of the provision and due to concerns on
whether financial and economic rights of women would be safeguarded let alone be advanced
. Besides HMA and SMA, marital disputes also involve laws meant to protect women from
domestic violence, dowry harassment and cruelty. The laws, paradoxically, are seen by the fledgling
men‘s rights movement as weakening the traditional family and contributing to marital discord (Basu
2006; Chowdhury 2014). These laws have implications on both marital stability and resolution of
The amendment lapsed following the dissolution of the parliament in 2014. The new government under the
Hindu-Nationalist BJP has said that it will not table the bill until it has examined the “grave and far reaching legal,
social and economic implications” of the amendment, including whether it will lead to increase in non-marriage, and
whether “there will be increase in illegitimate and live-in relations thereby fall of institution of marriage and family
values” (D. V. Sadananda Gowda‟s, Minister of Law and Justice, reply to Starred Parliament Question 373 in the
Lok Sabha on 18th December 2014).
In light of the complexities and difficulties in using legal venues to dissolve a marriage,
customary forms that emphasize, as Menski (2012) notes, interconnected social relationships rather
than individual rights are seen as an alternative. These customary norms might allow women to
claim agency on the basis of custom and open up new possibilities to maximize their rights in ending
a marriage (Holden 2008; Solanki 2011). While allowing some women to negotiate their right to end
marriage in terms of custom, they also subject many women to ―social boycott, ―persuasion,‖
shame, reprimand, and ridicule by caste and sect authorities‖ (Solanki 2011: 57). The possibilities
offered by communal and customary avenues are not ideal in a context where underlying social
relationships are unequal and power is gendered.
One of this study‘s contributions is the presentation of numbers to quantify prevalence and
variations in divorce and separation in India. These numbers move the discussion on divorce in
India from the realm of, what Parry (2001) termed as, ―quantitative gossip‖ to a much firmer
footing. The numbers show a rising trend of marital dissolution due to divorce or separation. The
overall level of marital dissolution, though, is still low. The findings also revealed important
differences by region and religion. The regional differences are attributable to the different kinship,
cultural and marriage systems. The religious differences are underpinned by both legal and cultural
norms of the respective communities. The findings also contribute to our understanding of the role
of education on martial stability. In India education promotes marital stability. As discussed above,
education might provide women with greater say in marriage, higher bargaining power and greater
gender equity which all might contribute to stronger marital bonds.
The study‘s findings were based on survey data that recoded marital status information. A
limitation of such data is that it views marital dissolution as an event. In India marital dissolution
lacks the concreteness of a definite break granted by a legal authority. Instead marital dissolution is a
process that is fluid and ambiguous with blurred boundaries. Therefore, there is a need for collection
of data using survey instruments that are responsive and sensitive to the complexities and pluralities
of marital dissolution in India. Dissolution of conjugal bonds, as described in a different Asian
context by Brickell and Platt (2013) but equally applicable to India, are messy, heterogeneous, non-
linear with creative construction of marital life course. To build a nuanced understanding would
require a move away from simple recording of current marital status, as in almost all major
demographic surveys in India, and to pay more attention to marital histories and biographies of both
men and women.
Abraham, Janaki. 2014. ―Contingent caste endogamy and patriarchy,‖ Economic and Political Weekly
Abraham, Leena. 2001. ―Redrawing the Lakshman Rekha: Gender differences and cultural
constructions in youth sexuality in urban India,‖ South Asia: Journal of South Asian Studies
Abeyasekera, Asha L. 2013. The Choosing Person: Marriage, Middle-Class Identities, and Modernity in
Contemporary Sri Lanka. PhD dissertation, Department of Social and Policy Sciences:
University of Bath.
Agnes, Flavia. 2009. ―Conjugality, property, morality and maintenance,‖ Economic and Political Weekly
Ahearn, Laura M. 2001. Invitations to Love: Literacy, Love Letters, and Social Change in Nepal. Ann Arbor:
The University of Michigan Press.
Allendorf, Keera. 2012. ―Marital quality from a rural Indian context in comparative perspective,‖
Journal of Comparative Family Studies 43(4): 527-544.
——. 2013. ―Schemas of marital change: From arranged marriages to eloping for love,‖ Journal of
Marriage and Family 75(2): 453-469.
Allison, Paul D. 2012. Survival Analysis Using SAS: A Practical Guide. Second Edition. Cary, NC: SAS
Andrist, Lester, Manjistha Banerji and Sonalde Desai. 2013. ―Marital negotiations: Examining the
gap between marriage and cohabitation in India,‖ in R. Kaur and R. Palriwala (eds.), Marrying
in South Asia: Shifting Concepts, Changing Practices in a Globalising World. New Delhi: Orient
Blackswan, pp. 116-140.
Aura, Siru. 2008. Women and Marital Breakdown in South India: Reconstructing Homes, Bonds and Persons.
PhD dissertation, Anthropology: University of Helsinki.
BBC. 2015. Party girls and too much sex: Why some Indian men seek divorce [5th August 2015].
Baas, Michiel. 2009. ―The IT caste: Love and arranged marriages in the IT industry of Bangalore,‖
South Asia: Journal of South Asian Studies 32(2): 285-307.
Basu, Srimati. 2006. ―Playing off courts: The negotiation of divorce and violence in plural legal
settings in Kolkata,‖ The Journal of Legal Pluralism and Unofficial Law 38(52):41-75.
——. 2012. ―Judges of normality: Mediating marriage in the family courts of Kolkata, India,‖ Signs
Berti, Daniela. 2011. ―Courts of law and legal practice,‖ in I. Clark-Decès (ed.), A Companion to the
Anthropology of India. Oxford, UK: Wiley-Blackwell, pp. 355-370.
Bharadwaj, Aditya. 2003. ―Why adoption is not an option in India: The visibility of infertility, the
secrecy of donor insemination, and other cultural complexities,‖ Social Science & Medicine
Bongaarts, John and Christophe Z. Guilmoto. 2015. ―How many more missing women? Excess
female mortality and prenatal sex selection, 1970–2050,‖ Population and Development Review
Bose, Sunita and Scott J. South. 2003. ―Sex composition of children and marital disruption in India,‖
Journal of Marriage and Family 65(4): 996-1006.
Brickell, Katherine and Maria Platt. 2013. ―Everyday politics of (in)formal marital dissolution in
Cambodia and Indonesia,‖ Ethnos 80(3): 293-319.
Chakraborty, Kabita. 2012. ―Virtual mate-seeking in the urban slums of Kolkata, India,‖ South Asian
Popular Culture 10(2): 197-216.
Chakraborty, Tanika and Sukkoo Kim. 2010. "Kinship institutions and sex ratios in India,"
Demography 47(4): 989-1012.
Chakravarti, Uma. 1993. ―Conceptualising brahmanical patriarchy in early India: Gender, caste, class
and state,‖ Economic & Political Weekly 27(14): 579-585.
Chen, Wan-Chi. 2012. ―The changing pattern of educational differentials in divorce in the context of
gender egalitarianization: The case of Taiwan,‖ Population Research and Policy Review 31(6):831-
Choudhury, Cyra Akila. 2008. ―(Mis) appropriated liberty: Identity, gender justice, and Muslim
Personal Law reform in India,‖ Columbia Journal of Gender & Law 17(1): 45-110.
Chowdhry, Prem. 2004. ―Caste Panchayats and the policing of marriage in Haryana: Enforcing
kinship and territorial exogamy,‖ Contributions to Indian sociology 38(1&2):1-42.
Chowdhury Romit. 2014. ―Family, femininity, feminism: ‗Structures of feeling‘ in the articulation of
men‘s rights,‖ in K.B. Nielsen and A. Waldrop (eds.) Women, Gender, and Everyday Social
Transformation in India. London: Anthem Press, pp. 189–202.
Das Gupta, Monica. 2010. ―Family systems, political systems and Asia's ‗missing girls,‖ Asian
Population Studies 6(2): 123-152.
de Graaf, Paul M. and Matthijs Kalmijn. 2006. ―Change and stability in the social determinants of
divorce: A comparison of marriage cohorts in the Netherlands,‖ European Sociological Review
de Maaker, Erik and Vibha Joshi. 2007. ―The northeast and beyond: Region and culture,‖ South Asia:
Journal of South Asian Studies 30(3): 381-390.
Desai, Sonalde and Lester Andrist. 2010. ―Gender scripts and age at marriage in India,‖ Demography
Desai, Sonalde and Gheda Temsah. 2014. ―Muslim and Hindu women‘s public and private
behaviors: Gender, family, and communalized politics in India,‖ Demography 51(6): 2307-
Dikshit, K. R. and Jutta K. Dikshit. 2014. North-East India: Land, People and Economy. New York:
Dommaraju, Premchand. 2009. ―Timing of first birth in India,‖ Genus 65(2): 81-101.
Dommaraju, Premchand and Victor Agadjanian. 2009. ―India‘s North–South divide and theories of
fertility change,‖ Journal of Population Research 26(3): 249-72.
Dommaraju, Premchand and Gavin Jones. 2011. ―Divorce trends in Asia,‖ Asian Journal of Social
Science 39(6): 725-50.
Dube, Leela. 2001. Anthropological Explorations in Gender: Intersecting Fields. New Delhi: Sage
Dutta, Diya, 2011. ―At the crossroads? Gender and multiple patriarchies in marriage and family in
contemporary Bengali society in Kolkata,‖ in S. Samita, R. Biswas and N. Dhawan (eds.),
Intimate Others: Marriage and Sexualities in India. Kolkata: Stree, pp. 173-194.
Fuller, C.J. and Haripriya Narasimhan. 2008. "Companionate marriage in India: The changing
marriage system in a middle-class Brahman subcaste," Journal of the Royal Anthropological
Garg, Sampak P. 1998. ―Law and religion: The divorce systems of India,‖ Tulsa Journal of Comparative
and International Law 6(1):1-20.
Ghosh, Partha S. 2009. ―Politics of personal law in India: The Hindu-Muslim dichotomy,‖ South
Asia Research 29(1):1-17.
Gilbertson, Amanda. 2014. ―From respect to friendship? Companionate marriage and conjugal
power negotiation in middle-class Hyderabad,‖ South Asia: Journal of South Asian Studies 37(2):
Greenberg, Judith G. 2003. ―Criminalizing dowry deaths: The Indian experience,‖ American University
Journal of Gender Social Policy and Law 11(2): 801-845.
Grover, Shalini. 2009. ―Lived experiences: Marriage, notions of love, and kinship support amongst
poor women in Delhi,‖ Contributions to Indian sociology 43(1): 1-33.
——. 2011. ― 'Purani aur nai shaadi': Separation, divorce, and remarriage in the lives of the urban
poor in New Delhi,‖ Asian Journal of Women's Studies 17(1): 67-99.
Guilmoto, Christophe Z. 2008. ―Economic, social and spatial dimensions of India's excess child
masculinity,‖ Population(E) 63(1): 91-118.
Härkönen, Juho and Jaap Dronkers. 2006. ―Stability and change in the educational gradient of
divorce. A comparison of seventeen countries,‖ European Sociological Review 22 (5): 501-517.
Harlan, Lindsey and Paul B. Courtright. 1995. ―Introduction: On Hindu marriage and its margins,‖
in L. Harlan and P.B. Courtright (eds.), From the Margins of Hindu Marriage : Essays on Gender,
Religion, and Culture. New York: Oxford University Press, pp. 3-18.
Holden, Livia. 2008. Hindu Divorce: A Legal Anthropology. Aldershot: Ashgate.
Hornbeck, Amy, Bethany Johnson, Michelle LaGrotta and Kellie Sellman. 2007. ―The protection off
women from domestic violence act: Solution or mere paper tiger?,‖ Loyola University Chicago
International Law Review 4(2): 273-307.
International Institute for Population Sciences ([IIPS]. 2010. District Level Household and Facility Survey,
2007-08. Mumbai, India: IIPS.
International Institute for Population Sciences [IIPS] and Population Council. 2010. Youth in India:
Situation and Needs 2006. Mumbai: IIPS.
Jain, Pratibha. 2005. ―Balancing minority rights and gender justice: The impact of protecting
multiculturalism on women's rights in India,‖ Berkeley Journal of International Law 23(1): 201-
Jaiswal, Tulika, 2014. Indian Arranged Marriages: A Social Psychological Perspective. New York: Routledge.
Jauregui, Beatrice and Tara McGuinness. 2003. ―Inter-community marriage and social change in
contemporary India: Hybridity, selectivity and transnational flows,‖ South Asia: Journal of
South Asian Studies 26(1): 71-85.
Jeffery, Patricia. 2014. ―Supply-and-demand demographics: Dowry, daughter aversion and marriage
markets in contemporary north India,‖ Contemporary South Asia 22(2):171-188.
Jennings, Elyse. 2014. ―Marital discord and subsequent dissolution: Perceptions of Nepalese wives
and husbands,‖ Journal of Marriage and Family 76(3):476-488.
Jesmin, Sonia and Sarah Salway. 2000. ―Marriage among the urban poor of Dhaka: Instability and
uncertainty,‖ Journal of International Development 12(5): 689–705.
Jones, Justin. 2010. ― ‗Signs of churning‘: Muslim Personal Law and public contestation in twenty-
first century India,‖ Modern Asian Studies 44(1): 175-200.
Kalpagam, U. 2008. ―Marriage norms, choice and aspirations of rural women,‖ Economic and Political
Weekly 43(21): 53-63.
Kaur, Ravinder and Priti Dhanda. 2013. ―Surfing for spouses: Marriage websites and the ‗new‘
Indian marriage?,‖ in R. Kaur and R. Palriwala (eds.), Marrying in South Asia: Shifting
Concepts, Changing Practices in a Globalising World. New Delhi: Orient Blackswan, pp.
Kishwar, Madhu. 1999. ―Love and marriage,‖ in M. Kishwar and R. Vanita (eds.), Off the Beaten
Track: Rethinking Gender Justice for Indian Women. Delhi: Oxford University Press, pp. 192–208.
Kodoth, Praveena. 2008. ―Gender, caste and matchmaking in Kerala: A rationale for dowry,‖
Development and Change 39(2): 263-283.
Kohli, Hari Dev. 2010. Supreme Court on Hindu Law. Delhi: Universal Law Publishing Co. Pvt. Ltd.
Kreager, Derek A., Richard B. Felson, Cody Warner and Marin R. Wenger. 2013. ―Women‘s
education, marital violence, and divorce: A social exchange perspective,‖ Journal of Marriage
and Family 75(3): 565-581.
Law Commission of India. 1978. The Hindu Marriage Act, 1955- Irretrievable Breakdown of Marriage as a
Ground of Divorce. Report no: 71. New Delhi: Government of India.
——. 1983. The Grounds of Divorce Amongst Christians in India: Section 10 Indian Divorce Act 1869. Report
no: 90. New Delhi: Government of India.
——. 2008a. Laws on Registration of Marriage and Divorce – A Proposal for Consolidation and
Reform. Report no. 211. New Delhi: Government of India.
——. 2008b. Laws of Civil Marriages in India –A Proposal to Resolve Certain Conflicts. Report no. 212.
New Delhi: Government of India.
——. 2009. Irretrievable Breakdown of Marriage –Another Ground for Divorce. Report no. 217. New Delhi:
Government of India.
Lee, Yean-Ju . 2006. ―Risk factors in the rapidly rising incidence of divorce in Korea,‖ Asian
Population Studies 2(2):113–131.
Lemons, Katherine. 2010. At the Margins of Law: Adjudicating Muslim Families in Contemporary Delhi.
PhD dissertation, Anthropology: University of California, Berkeley.
——. 2013. ―When marriage breaks down how do contracts matter? Marriage contracts and divorce
in contemporary north India,‖ in R. Kaur and R. Palriwala (eds.), Marrying in South Asia:
Shifting Concepts, Changing Practices in a Globalising World. New Delhi: Orient
Blackswan, pp. 371-388.
Lu, Jiehua and Xiaofei Wang. 2014. ―Changing patterns of marriage and divorce in today‘s China,‖
in I. Attané and B. Gu (eds.), Analysing China's Population: Social Change in a New Demographic
Era. London: Springer, pp. 37-49.
Marak, Quinbala. 2012. ―A note on kinship studies in North-East India,‖ South Asian Anthropologist
Matysiak, Anna, Marta Styrc and Daniele Vignoli. 2014. ―The educational gradient in marital
disruption: A meta-analysis of European research findings,‖ Population Studies 68(2): 197-215.
Mehta, Bhamini and Shagufa Kapadia. 2008. ―Experiences of childlessness in an Indian context: A
gender perspective,‖ Indian Journal of Gender Studies 15(3): 437-460.
Menski, Werner F. 2001. Modern Indian Family Law. Richmond: Curzon.
——. 2012. ―Ancient and modern boundary: Crossings between personal laws and civil law in
composite India,‖ in J.A. Nichols (ed.), Marriage and Divorce in a Multicultural Context : Multi-
Tiered Marriage and the Boundaries of Civil Law and Religion. New York: Cambridge University
Press, pp. 219-252.
Mukhopadhyay, Madhurima. 2011. ―Choice an agency in marital decisions: A study among Hindu
Bengalis across class in Kolkata,‖ in S. Samita, R. Biswas and N. Dhawan (eds.), Intimate
Others: Marriage and Sexualities in India. Kolkata: Stree, pp. 121-148.
Nandy, Amrita. 2015. ―Natural mother = real mother? Choice and agency among un/natural
‗mothers‘ in India,‖ Women's Studies International Forum 53: 129-138.
Netting, Nancy S. 2010. ―Marital ideoscapes in 21st-century India: Creative combinations of love
and responsibility,‖ Journal of Family Issues 31(6): 707-726.
Nichols, Joel A. 2012. ―Multi-tiered marriage: Reconsidering the boundaries of civil law and
religion,‖ in J.A. Nichols (ed.), Marriage and Divorce in a Multicultural Context : Multi-
Tiered Marriage and the Boundaries of Civil Law and Religion. New York: Cambridge
University Press, pp.11-59.
Osella, Caroline and Filippo Osella. 2006. Men and Masculinities in South India. London: Anthem Press.
Palit, Manjushree. 2014. Uncovering Messages of Intimacy in Urban, Educated, and Middle-Income Married
Indian Couples: A Phenomenological Study Using Pictures as Metaphors. PhD dissertation, Human
Development, Virginia Polytechnic Institute and State University, Blacksburg, VA.
Palriwala, Rajni. 1991. ―Transitory residence and invisible work: Case study of a Rajasthan village,‖
Economic and Political Weekly 26(48): 2763-2772.
Park, Hyunjoon and James M. Raymo. 2013. ―Divorce in Korea: Trends and educational
differentials,‖ Journal of Marriage and Family 75(1):110-126.
Parliament of India. 2011. Forty Fifth Report On The Marriage Laws (Amendment) Bill, 2010. New Delhi:
Secretariat Department Related Parliamentary Standing Committee On Personnel, Public
Grievances, Law And Justice.
Parry, Jonathan P. 2001. ―Ankalu's errant wife: Sex, marriage and industry in contemporary
Chhattisgarh,‖ Modern Asian Studies 35(4):783-820.
——.2004. ―The marital history of a ‗thumb impression man‘ ‖ in D. Arnold and S. Blackburn
(eds.), Telling lives: South Asian life histories. New Delhi/Bloomington: Permanent
Black/Indiana University Press, pp. 281-318.
Prakash, Ravi and Abhishek Singh. 2013. ―Who marries whom? Changing mate selection
preferences in urban India and emerging implications on social institutions,‖ Population
Research and Policy Review 33(2): 205-227.
Redding, Jeffrey A. 2012. ―What American legal theory might learn from Islamic law: Some lessons
about 'the rule of law' from 'Shari‗a Court' practice in India,‖ University of Colorado Law Review
Riessman, Catherine Kohler. 2000. ―Stigma and everyday resistance practices: Childless women in
south India,‖ Gender & Society 14(1):111-135.
Singer, Judith D. and John B. Willett. 2003. Applied Longitudinal Data Analysis: Modeling Change and
Event Occurrence. New York: Oxford University Press.
Singh, Kirti. 2013. Separated and Divorced Women in India: Economic Rights and Entitlements. New Delhi
and Ottawa: Sage Publications and International Development Research Centre.
Singh, Ajay Kumar and R. K. Sinha. 2005. ―Growing incidence of divorce in Indian cities: a study of
Mumbai,‖ Paper Presented at the XXV IUSSP International Population Conference, Tours,
Solanki, Gopika. 2011. Adjudication in Religious Family Laws: Cultural Accommodation, Legal Pluralism, and
Gender Equality in India. Cambridge, U.K.: Cambridge University Press.
Still, Clarinda. 2010. ―Spoiled brides and the fear of education: Honour and social mobility among
Dalits in south India,‖ Modern Asian Studies 45(5): 1119-1146.
Subramanian, Narendra. 2008. ―Legal change and gender inequality: Changes in Muslim family law
in India,‖ Law & Social Inquiry 33(3): 631-672.
Titzmann, Fritzi-Marie. 2013. ―Changing patterns of matchmaking: The Indian online matrimonial
market,‖ Asian Journal of Women's Studies 19(4): 64-94.
United Nations. 1999. Demographic Yearbook: Historical Supplement 1948-1997. New York: United
Vatuk, Sylvia. 2013a. ―The application of Muslim Personal Law in India: A system of legal pluralism
in action,‖ in E. Giunchi (ed.), Adjudicating Family Law in Muslim Courts. London: Routledge,
——. 2013b. ―The ‗Women's Court‘ in India: An alternative dispute resolution body for women in
distress,‖ The Journal of Legal Pluralism and Unofficial Law 45(1): 76-103.
Walters, Kimberly, Rakhi Dandona , Lawrence C. Walters , Vemu Lakshmi , Lalit Dandona and
John A. Schneider. 2012. ―Wives without husbands: Gendered vulnerability to sexually
transmitted infections among previously married women in India,‖ AIDS Care: Psychological
and Socio-medical Aspects of AIDS/HIV 24(9): 1103-1110.
White, Sarah C. 2013. ―Patriarchal investments: marriage, dowry and economic change in rural
Bangladesh,‖ University of Bath: Centre for Development Studies. Bath Papers in
International Development, no. 19.
Xaxa, Virginius. 2004. ―Women and gender in the study of tribes in India,‖ Journal of Gender Studies
Figure 1: Per cent divorced or separated among ever married age 15 and over, by
age group and year
Source: For 1961, 1971, 1981 and 1991, Indian census data as reported in United Nations (1999). For
2001 and 2011, Indian census data from the tables published by the Office of the Registrar General &
Census Commissioner of India, various years.
Figure 2: Probability of divorce or separation for women, by year, with 95 per cent
confidence intervals (shaded)
Source: Author’s calculation based on DLHS-3 data.
Figure 3: Per cent of marriages ending in divorce or separation, women, survival
(Figure 3 continues in the next page)
Figure 3 continued
C: Education in years
Source: Author’s calculation based on DLHS-3 data.
Table 1: Sample characteristics, per cent or mean
Education, women, (%)
Less than 5 years
5 to 9 years
10 or more years
Education, husband, (%)
Less than 5 years
5 to 9 years
10 or more years
Marriage cohort, (%)
Living children, (%)
At least one son
Other Backward Class
Marriage age, mean
Marital duration, mean
Number of respondents
Source: Author’s calculation based on DLHS-3 data.
Table 2: Per cent and number divorced and separated among ever married age 15
and over, by year
Per cent divorced or separated among ever
married 15 & over
Number divorced or
separated 15 & over
Source: For 1961, 1971, 1981 and 1991, Indian census data as reported in United Nations (1999). For
2001 and 2011, Indian census data from the tables published by Office of the Registrar General &
Census Commissioner of India, various years.
Table 3: Odds ratio of divorce or separation, unadjusted and adjusted estimates
Less than 5 years
5 to 9 years
10 or more years (ref.)
Less than 5 years
5 to 9 years
10 or more years (ref.)
At least one son (ref.)
Other Backward Class
Marital duration squared
Notes: a: Marital duration is controlled in all the models.
Significance levels: ** p < 0.01; * p < 0.05.
Source: Author’s analyses based on DLHS-3 data.
Table 4: Odds ratio of education and marriage cohort, interaction model
No or lower vs.
Middle vs. higher
No or lower vs.
Note: The model controls for all covariates as in Table 3.
Source: Author’s analyses based on DLHS-3 data.