Abstract
In India marrying across the caste kindles strong community resentment, leading to the extent of honor killing, yet few couples dare to defy this stringent social norm. Analysis of large-scale survey (India Human Development Survey 2011–12) data exhibits an inconsequential rise in this social incongruity since 1951 to attain 4.5% in 2012. It is most prevalent in the northeastern region (11.6%), but least prevalent in caste-ridden central India (1.8%). Multi-variate statistics exhibit that if women are allowed to choose their life-partner, caste takes a rear-seat in marriage contemplation. It is quite prominent among the women who selected the husband by themselves and knew the husband at least one year before the marriage. Contrary to general notion, education is not able to promote inter-caste marriage. Odds of inter-caste marriage taking place in Dalit (lowest social standing) households is much lesser than the higher caste. Resilient targeted efforts are necessary to promote inter-caste marriage, which may loosen the noose of the caste system in India.
Introduction
Marriage is almost universal in India, and there are various crucial considerations for mate selection. Some of the most important ones are caste, religion, mother tongue, economic condition, etc. Among them, caste similarity between the two uniting families is the foremost important consideration. Such similarities are regarded as the backbone of a successful marriage. Caste, unlike the class, is prescribed and not ascribed. It is pre-decided at the time of birth itself, and one cannot change it throughout life. It is also embedded in the Hindu religion. Hence, it is deeply rooted in Indian culture and society. Compounding the caste interplay, monogamy is the prevalent form of marriage and divorce not readily acceptable to the majority of communities. So, it is quite essential to choose a life-partner carefully. As a result, inter-caste and inter-religious marriages are socially condemned and frowned upon (Srinivasan and James, 2015). Hence, even today, mostly arranged marriage is preferred over the other types (Banerji et al., 2013; Desai and Andrist, 2010; Jeejeebhoy et al., 2013; Mishra, 2013), especially among the parents and elders. The older people prefer it to get social solidarity in hard times, whereas the younger generation may follow it to get parental support.
With the globalization and socio-economic development, social norms are supposed to loosen its noose. Increase in the level of education, change in place of residence, easy mingling with opposite sex, etc. opens the gateway for selecting a life-partner from another community. One of the famous Indian sociologists (Srinivas, 1955) noted back in 1955 that co-education is bound to make inter-caste marriages more common shortly. Further, he concluded that the caste system is bound to give way in the end, though it would be against the nature of things to expect untouchability to disappear overnight. A similar study (Kales, 2014) suggests that even today the caste system is prevalent mainly in rural India, which condemns inter-caste marriage, and the couple who tries to defy this norm faces dire consequences. Also, the sporadic incidences of honor killing, especially in caste-ridden north India, negate such inference. Many parents wish to send their children to modern schools and provide all modern facilities. But when it comes to mate selection, they would like to follow the traditional pattern of marrying within the same caste. Here the problem arises. Due to strong cultural importance put into the marriage and caste system embedded in the Hindu religion, inter-caste marriages (Baas, 2007) and marriages outside the same economic status groups are not common phenomena (Goli et al., 2013). Marriage outside the caste is considered taboo, although there are no restrictions on inter-caste marriage as per the Hindu Marriage Act (Trigunayat, 2014). Such a couple is out-casted, driven out of the parental home, and the worst cases lead to honor killing. At the backdrop of such socio-cultural settings, it is pertinent to know which type of female and male take the high risk of breaking this arduous social norm.
Although previous research exhibits existence of a preference for same-caste marriage in India until today (Banerjee et al., 2012; Dommaraju, 2016), traces of inter-caste marriage can be found far back in history (Banerjee et al., 2012; Chaudhry and Mohan, 2011). Cited reasons for (long distance) inter-caste marriage is the poverty and inability to pay the dowry demand at the native place (Chaudhry and Mohan, 2011: 317). Another study by Goli et al. (2013) concludes that education and economic status are two major factors contributing toward women’s choice and freedom in the selection of a spouse from outside their own socio-economic group. But another study by Das et al. (2011) found education (of both women and men) to have a negative effect on inter-caste marriage, though the economic condition has a positive impact on it. However, studies suggest that today in Indian society we can see inter-caste marriages. Though mostly it is part of the city/urban culture, and they constitute a minor proportion of the total number of marriages (Ahuja and Ostermann, 2016; Das et al., 2011; Goli et al., 2013; Trigunayat, 2014). It is said that those who defied inter-caste marriage by violating the social norm had to face the consequences in terms of violence, social boycott, family boycott, and death of the boys and girls (honor killing) (Das et al., 2011). Research suggests a higher prevalence of inter-caste marriage among lower caste women (Ahuja and Ostermann, 2016) because interest in inter-caste marriage is rooted in a desire for upward mobility and governed by the principle of exchange (Ahuja and Ostermann, 2016). It is also substantiated by the findings that lower caste people in rural North India evaluate their lives to be worse than higher caste people and this difference is not explained by poverty (Spears, 2016).
There is a shortage of literature on inter-caste marriage, possibly due to the lack of data, or the controversies surrounding it, though it is a very pertinent and complex social issue in the Indian context. There are scattered ethnographic studies that, by their very nature, are not generalizable. Studies are primarily conducted in pockets and with small sample sizes. Further, some of the works (Das et al., 2011; Goli et al., 2013; Jejeebhoy et al., 2013) that cover the entire country and have large-scale data are not free from limitations. These studies have mainly investigated the prevalence of inter-caste marriage across different socio-economic strata, and region, but they try to assess the factors responsible for inter-caste marriage, including those variables that actually do not have a corresponding effect on inter-caste marriage. Hence, the present paper is a humble attempt to address unexplored areas in this regard. The overall aim is to investigate the level and trends of inter-caste marriage from 1951 to 2012 and its spatial variation in prevalence, and its determinants.
Methodology
The present study used data from the India Human Development Survey 2011–12 (IHDS-II), a nationally representative, multi-topic survey of 42,152 households. The survey interviewed women who were interviewed in the previous IDHS (2005) survey, or ever-married women in the age group 15 years and over. It surveyed total 39,523 women within this age group. Hence, the year of marriage for these women dates between 1951 and 2012. Out of these many samples, 29,596 non-tribal Hindu women are retained. Because the caste system is a mostly Hindu-religion feature, it is not possible to say with certainty about its prevalence in other faiths. Similarly, tribal population is outside the caste system and is therefore outside the scope of this study. Further, 133 women whose information on inter-caste marriage is missing were removed from the analyses. Thus, the analyses are related to 29,463 women aged 15 years and over. As IHDS-II provides a nationally representative and large sample data-set, it allows a robust statistical exploration on one of the very critical social issues, such as inter-caste marriage, and draws meaningful conclusions. However, it is not possible to do a state-wide analysis because of the small sample size for some of the states. Nevertheless, an attempt is made to analyze at the regional level. The study first presents levels and trends of inter-caste marriage from 1951 to 2012 by using a line graph with a linear trend. Spatial variation in prevalence of inter-caste marriage is presented as the bi-variate table with the chi-square test. To see the factors associated with inter-caste marriage, the bi-variate cross-tabulation is done and chi-square tests performed. To find out the determinants of inter-caste marriage, binary logistic regression is run. The number of cases in different categories of some variables may not add up to the whole total, due to the missing cases. The entire statistical exercise is performed in the SPSS-20 software package.
About the variables
Dependent variable
The question related to inter-caste marriage was asked: “Is your husband’s family the same caste as your natal family?” The answers were pre-coded as Yes or No. If the answer to this question was “No”, then it was the inter-caste marriage. But there was no follow-up question to assess whether she married a man of a either a lower or upper caste compared to her natal family.
Independent variables
The question relating to the difference in economic condition between the natal family of the woman and husband was asked: “At the time of your marriage, if you compared the economic status of your natal family with your husband’s family, would you say your natal family was…” and the options were given as Same, Natal better off, or Natal worse off. This variable is used to assess whether the economic condition overweighs the caste conformity.
What is your childhood place of residence? The possible options of the answer to this question were: 1) Same Village/Town, 2) Another Village, 3) Another Town, and 4) Metro City. This variable has been recoded into three categories. The first two options are used as it is, whereas the third and fourth options are combined to form “Another town/metro.”
Education of women. The information on education is available in single years from 0 to 15, and women with above a bachelor degree education are assigned “16.” It may be noted here that under the Indian system of formal education, 15 years of schooling is required to complete a bachelor’s degree. It has been recoded into 0 (as Illiterate), 1 to 5 years, 6 to 10 years, and 11 years and above. This regrouping is done based on the frequency distribution and conforms to the Indian system of education. Hence, persons with “zero” years of education mean “illiterate,” 1–5 years means “Primary,” 6–10 means “Upper Primary & Secondary”, and 11 years and above implies “Higher secondary and above” level of education.
Age at marriage. Relating to the age at marriage, two questions were asked. First: “How old were you when you got married?” Second: “How old were you when you first started living with your husband (had Gauna)?” For the present work, we have used the later question, because some parts of an Indian marriage are held at a very tender age, but couples do not live together until Gauna is performed. This variable is recoded into two groups: women who married below 18 years of age and women who married at 18 years or older. In India 18 years is the legal age of marriage for women.
Acquaintance with husband. A question was asked relating to the acquaintance with the husband before the marriage: “How long had you known your husband before you married him?” Five options were given to the women: 1) On wedding/Gauna day only, 2) Less than one month, 3) More than one month but less than one year, 4) More than one year, 5) Since childhood. It has been used as it is given.
Growing up. We asked the question: “Did you grow up in the same village/town as your husband?” The pre-coded answers were “Yes” and “No.” This variable has been used as it is. It has been used to understand whether growing up together can nullify the strong caste affinity.
Mate selection. All the married women were asked about who selected their: “Who chose your husband?” The four pre-decided answers were: 1) Respondent herself, 2) Respondent & parents/other relatives together, 3) Parents/Other relatives alone, 4) Others. In this study, the first two responses have been used as it is, whereas the later two are combined.
Caste of the head of the household. The caste of the head of every household is asked. IHDS provided a variable called GROUPS, which is the combination of caste and religion. Though it provided data on seven categories, after filtering the data to suit the present study, only four hierarchical categories remained. They are Brahmin, Forward caste, Other backward caste, and Dalits.
Region. The entire country is divided into six geographically continuous regions, viz. North, Central, East, Northeast, West, and South. They are socio-culturally homogenous within and heterogeneous between. This classification is used by one of the large-scale surveys in India; that is, the National Family Health Survey and other various studies.
Place of residence. In the present study we used two types for the place of residence. First the place of residence given in two categories; that is, rural and urban. The second type if place is provided in four categories: metro, other urban areas, more developed village, and less developed village. They are used in the given format without any alteration. They are used to see whether there is any variation within different urban set-ups and rural areas, and to distinguish the clear-cut differences between rural and urban areas.
In the present study, only those variables which correspond to the time of marriage are used. However, education of women and caste of the household are used under the assumption that they have not changed after marriage. In India, it is not customary for women to continue formal education after the wedding. As such, the women must have attained the educational level before entering into marital life. Further, unlike class, caste is a rigid system, as the caste of the household into which women have married does not changed due to her entry into the family. Hence, education and caste of the household have a direct correspondence to, and implications on, the inter-caste marriage.
Results
Inter-caste marriage: levels and trends
IHDS-2011–12 data provides information on marriages taking place from 1951 to 2012. The survey data reveals three distinct periods of inter-caste marriage in India: 1951–1978, 1978–1994, and 1994–2012 (Figure 1). The first period shows an extremely erratic pattern, the percentage of inter-caste marriage varying between zero to as high as 20%. In the second period, the variation is somewhat insignificant. During this period the percentage of inter-caste marriage hovered around 4.7. After that, in the third period, the trend is again erratic. However, the fluctuation during the later period is smaller than the first period. In the third period, it varied between 2.5 to 5.2%. Due to the significant fluctuations in the percentage of inter-caste marriage by single year data, we fitted a linear best fit line. The linear pattern shows a steady rise in the inter-caste marriage to attain slightly above five % by the year 2012.

Levels and trends of inter-caste marriage in India from 1951 to 2012.
From the total sample who belong to Hindu religion and non-tribals, 4.5% were found to have married across another caste in India during 2011 to 2012 (Table 1). A study by Goli et al. (2013) based on the IHDS-2005 data estimated inter-caste marriage to be about 6.1% in 2005. Whereas, another research based on National Family Health Survey 2005–2006 shows the percentage of inter-caste marriages as about 10%, of which 4.97% of women married men of a lower caste, and another 4.95% of women married men of a higher caste (Das et al., 2011). The variations between the present study and the earlier study on IHDS-1 might be because of the differences in retaining the sample size, whereas the difference with later work might be due to the sampling design effect, as these two large-scale studies have pretty different sampling designs. Whatever may be the level of prevalence of inter-caste marriage, it is certain that amidst stringent social norm it exists in some extent, so further investigation is required.
Percentage distribution of women by inter-caste marriage in 2011–2012.
Spatial variation in prevalence of inter-caste marriage
India is a vast country with significant regional variation in all aspects of human civilization. As such, it is pertinent to also review the regional difference in the prevalence of inter-caste marriage. It is found (Table 2) that the inter-caste marriage is notably (11.6%) higher in the northeastern region of India, whereas it is minimal in central India. This pattern is very similar to the earlier study by Das et al. (2011). The regional variation in inter-caste marriage also reflects the existing social norm. The Khap panchayat is an active social institution in central India known to be against inter-caste marriage, as such inter-caste marriage is much lower in this region. On the other hand, northeast India is known to be more liberal, and the caste system is not strongly followed, possibly due to intermingling with a large proportion of the tribal population. Further, as expected, inter-caste marriage is at a minimum in less developed villages of India. This indicates the persistence of a stringent caste system in the less developed villages, for which women/men are afraid of crossing the accepted social norm. Further, arranged marriage is the norm in such areas, and the question of inter-caste marriage hardly arises. Surprisingly, compared to the women in the metros, the women in other urban areas are slightly more inclined toward inter-caste marriage. This may be due to the following two reasons. First, in metros, people are too engrossed with hectic daily work schedule and have minimal social interaction. Whatever interaction they have is also within their caste or class only. As a result, metro areas do not provide much scope for inter-caste marriage. Second, in small urban areas, people are comparatively less busy and have more social interactions across a range of population, which provides sufficient ground for inter-caste marriage. A significant rural and urban dichotomy is apparent in inter-caste marriage. Inter-caste marriage is much higher in urban areas than in metro areas.
Spatial variation in prevalence of inter-caste marriage in India during 1951 to 2012.
The scenario of inter-caste marriage
Analysis of the data shows (Table 3) that inter-caste marriage is comparatively higher for women who reported that the economic condition of their natal family was better-off than the husband’s family at the time of marriage. Though the variation is pretty negligible and not significant, it may suggest that the economic condition of a girl’s parents is capable of attracting a marriage alliance from other castes as well. In other words, improvement in financial status may be able to dilute the caste mechanism to a certain extent.
Percentage of women by inter-caste marriage and background characteristics at the time of marriage.
The statistics reveal that the percentage of inter-caste marriage is significantly higher among the women who knew their husband for more than a year before the wedding (12.1%). It is only 3% among the women who knew their husband since childhood. But somewhat contrasting to this, it is noted that among the women who grew up in the same village about 7% experienced inter-caste marriage, whereas it is only 4% for other groups. This indicates that inter-mingling with people of different castes people opens the vistas for inter-caste marriage. Growing up in the same village may give an opportunity for youngsters to meet in various platforms like school, marriage, religious ceremony, or marketplace, and may develop a liking for each other that ultimately leads to inter-caste marriage.
The results (Table 3) show that, when women selected husband by themselves, as high as one-fifth of them selected across the caste. This suggests that given a chance, the young people may not give the caste much weight in selecting their life-partner. The criteria for mate selection for them may be different from older people. Another study shows that, due to poverty and the inability to meet the dowry demand, parents wanted them (daughter) to marry across caste, so they had no choice (Chaudhry and Mohan, 2011: 317). A weak positive linkage between the level of education and inter-caste marriage is observed. Among the illiterate women, about 4% married across the caste, whereas it is about 5% among women who are matriculate or above. However, the variation is very marginal and statistically not significant. Another relevant variable that may affect the inter-caste marriage is age at marriage. The present study reveals that younger women are less likely to go for inter-caste marriage, whereas it is opposite for women who married after 18 years of age. Usually, women who married at a younger age have a lower level of education and less decision-making power, so parents marry them within the same caste. Women who marry at older ages are more likely to be better educated, and thereby have more decision-making power concerning mate selection. Further, data reveals that there is a somewhat negative relationship between social-hierarchy and inter-caste marriage. It is lowest in the Dalit households, with the maximum observed in Brahmin households. However, the variation is very marginal and statistically not significant to draw a strong conclusion. This pattern observed in the present work is similar to an earlier study by Ahuja and Ostermann (2016).
Determinants of inter-caste marriage
The multi-variate statistical results (Table 4) show that the economic condition of the natal family plays a significant, positive role in inter-caste marriage. Compared to the same economic condition, if the economic condition of the natal family was better than the husband’s family, the odds of getting married across the caste is comparatively higher (odds ratio 1.209). Compared to women who grew up with their husband in the same village, women growing up in other villages are less likely to marry across the caste (odds ratio: 0.646). In other words, the chances of marrying across the caste are high if they grow up in the same village or town. One of the notable findings of the present study is that education is not able to promote inter-caste marriage in India. There is a clear negative relationship between education of women and inter-caste marriage. Compared to the illiterate, the chances of marrying across the caste reduces (odds ratio: 0.808–0.793) with the increase in the level of education. This finding is similar to one of the large-scale studies (Das et al., 2011), but contrary to another study (Goli et al., 2013).
Results of binary logistic regression.
Dependent variable: Marriage type: Same caste=0, Inter-caste=1.
The odds of getting married across the caste is slightly more among women who married when above the legal age. Compared to the women who met her husband for the first time around the day of the wedding, an acquaintance for less than a year is less likely to move toward inter-caste marriage. But women who knew their husband for more than a year are more likely to experience inter-caste marriage. This may mean that, to lead to inter-caste marriage, having an acquaintance or some relation of at least a year is necessary. But contrary to this, knowing the husband since childhood does not favor inter-caste marriage, because the woman is likely to be aware of everything regarding the prospective husband.
One of the most notable findings of the present study is that the women’s decision in selecting the husband has a significant role in inter-caste marriage. If women are allowed to choose their life-partner, caste takes a rear-seat in marriage contemplation. On the other hand, if the husband is selected by parents or relatives, caste is a significant consideration and there is no room for inter-caste marriage. Compared to the Brahmins, the odds of inter-caste marriage taking place in the Dalit households are much less (OR: 0.733, CI: 0.579–0.930). This finding is similar to one of the earlier studies by Ahuja and Ostermann (2016). This result implies that women from a lower caste are marrying into the Brahmin households. It may be possible because the educated Dalit women may be acceptable to the Brahmin men, as they appear to be on par with other upper caste women, except regarding caste. On the other hand, a Dalit women marrying higher caste men may be one of the ways of climbing the rigid caste hierarchy. If not for herself, at least for her future offspring, which is otherwise not possible.
Discussion
Inter-caste marriage is an antidote to the caste system, one of the most deeply rooted social institutions in India. It is believed that modernization, education, exposure to mass media, inter-mingling, etc. may lay a favorable foundation for inter-caste marriage. In this context, it is pertinent to mention here that the female literacy rate in India has risen from a mere 7.93 in 1951 to 64.6 in 2011, and male literacy from 16.67 to 80.9. Further, the educational level has also increased greatly. According to Census 2011, about 47.2% of the households had television, 63% had mobile/telephone connectivity, and 21% had a scooter/moped/motorcycle. Further, 3.1% had a computer or laptop with internet connection and 6.3% without internet. Still, the rise in the inter-caste marriage during 1951 to 2012 is not very encouraging, because it is true that it would go against nature to expect untouchability (caste system) to disappear overnight (Srinivas, 1955). Also, during the last six decades, the current of modernization and Westernization are much stronger, but the inter-caste marriage is not increasing as expected, despite such favorable atmospheres. So, it may also mean that, though the attitude of the people may have changed to some extent, it is not sufficient enough to transform to action. Education is considered to be the panacea of all the social evils, and it is expected that, with the increase in the level of education, inter-caste marriage also will increase. But the data reveals a different picture in the present study, as well as in another study (Das et al., 2011). The empowerment in choosing the husband and sufficient acquaintance with husband before marriage has a profound impact on inter-caste marriage. This indicates that educated women prefer to marry within the same caste, though they may not be against the inter-caste marriage. Further, a very low prevalence of inter-caste marriage in India corroborates the acceptance of the caste system. The findings of the present study are also pre-cursory toward a need for a dedicated, large-scale study on inter-caste marriage.
Conclusions
The negative attitudes toward inter-caste marriage is a robust indicator for assessing social distance across the different groups, whereas a positive attitude is a definite step toward reducing caste hierarchy, thereby leading to the caste-assimilation. Further, inter-caste marriage is the best remedy to annihilate this deep-rooted social menace. From the ongoing study, it is clear that the traces of inter-caste marriage can be found over a long period. But its trends and the current level are not very encouraging. In contrast to the common belief and expectation, education does not seem to promote inter-caste marriage; instead, empowering women to choose her life-partner will likely boost inter-caste marriage. There is a need for resilient targeted efforts to encourage inter-caste marriage so that the noose of the caste system in India can be loosened.
Footnotes
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
