Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 172
The Changing Transitions to Adulthood in Developing Countries: Selected Studies 6 Marriage Patterns in Rural India: Influence of Sociocultural Context Shireen J. Jejeebhoy and Shiva S. Halli There is considerable agreement that notable changes have occurred in India in the timing of marriage. For example, the singulate mean age at marriage of females increased from 15.2 in 1951 to 19.0 by 1991 (Bhat and Halli, 1999) and to 19.7 by 1998-1999 (IIPS and ORC Macro, 2000). However, regional variation is quite evident. For example, in 1992-1993 the median age at first marriage was 15.1 among women ages 25 to 49 in Uttar Pradesh in north India, compared to 18.1 in Tamil Nadu in south India (IIPS, 1995). Less is known about regional differences and trends in marriage patterns, such as endogamy, postmarital residence patterns, spousal age and educational differences, dowry, and the extent to which women have a say in determining timing and partner, on the one hand, and the disbursal of their dowries on the other. Also poorly understood is the extent to which changes in these patterns are conditioned by sociocultural factors such as region and religion and their association with female autonomy. A unique data set provides an opportunity to explore marriage patterns and differences among successive cross-sections of Hindu and Muslim women who were married in the roughly 25 years from 1968 to 1993 in two socioculturally heterogeneous settings, namely rural areas of Uttar Pradesh and Tamil Nadu. The intention of this chapter is to describe differences in marital age and patterns among successive marital age cohorts and explore the extent to which differences emerge by region and religion. Many arguments have been postulated to explain increases in marital age. Some would attribute the increase as a response to a marriage squeeze rather than as an outcome of increased educational attainment, which they argue is not advanced enough to be incompatible with early marriage (Bhat
OCR for page 173
The Changing Transitions to Adulthood in Developing Countries: Selected Studies and Halli, 1999). They argue that India has begun to face an excess supply of women of marriageable ages because of changes such as declining infant and child mortality and the reduction in numbers of widowers available as maternal mortality declined, as well as because women tend to marry men who are older than them. This view would argue that this change resulted in both a longer search for a suitable husband and higher dowries (Amin and Cain, 1997; Bhat and Halli, 1999; Caldwell, Reddy, and Caldwell, 1983; Rao, 1993). Others would argue that increases in marital age may be attributed to shifts in the education of boys and girls and the imposition of a legal minimum age at marriage (Amin and Cain, 1997). It is well known that marriage patterns reflect a fundamental difference between women from north and south India, and Hindu and Muslim women (see, e.g., Dyson and Moore, 1983; Karve, 1965). However, the extent to which these patterns are changing over time is less well studied. For example, there has been considerable public education on the problems associated with early marriage and laws against marriage to females under 19 and these, along with a growing recognition of the importance of educating females, would argue for an increase in marital age. Legal sanctions exist, at least in theory, against the practice of dowry; yet the consumer culture and the greater education levels of young males are argued to have prompted families to demand larger dowries than before. In the more patriarchal kinship structure prevailing in the north, and particularly among the Hindus, marriage is regarded as an alliance of two families and involves the incorporation of outsiders as wives into the family. The resulting village exogamy prevailing in the north ensures a break between the natal family and the family into which a woman is married: not only is a woman married off into a distant village, but kinship rules ensure that, by and large, no other women from her natal family, or even village, can be married into the same village (Karve, 1965). The practice of marrying young girls into distant villages and into families with which previous contact has been limited at best and subsequent contacts are usually infrequent heightens women’s powerlessness (Committee on the Status of Women in India, 1973). In contrast, north Indian Muslims are much more likely than Hindus to marry kin, and less likely to practice village exogamy. In the south, both Hindu and Muslim women enjoy less alienating marriage ties. Here, marriage is more a means of consolidating existing kinship networks than a political alliance. As a result, there are fewer restrictions on marriage within the village or within easy travelling distance from the woman’s natal village. In Tamil Nadu, marriages often take place among affines. As in the north, the practice of dowry is common. Although in the north the pattern and flow of resources is strictly one way, even after marriage (Das Gupta, 1987), in Tamil Nadu, women themselves appear to have more control over their dowries. Unlike women in the north who are traditionally per-
OCR for page 174
The Changing Transitions to Adulthood in Developing Countries: Selected Studies ceived as temporary members in their natal homes (Dube, 1988), ties between women and their natal families in south India remain close even after marriage; while daughters may continue to receive gifts from their parents and nearest kin even after marriage (see, e.g., Bhat and Halli, 1999), in turn, the daughters provide emotional and sometimes (less frequently) economic support to their parents. This chapter draws on data from a community-based survey on women’s autonomy in two culturally distinct sites: Uttar Pradesh in the north, in which the situation of women is especially poor, and Tamil Nadu in the south, where gender relations are somewhat more balanced, and women are relatively better off. About half the sample in each setting was composed of Hindu women, the other half Muslim women. The survey was conducted in 1993-1994 and included questionnaires for married women ages 15 to 39 and their husbands, if available. Findings presented elsewhere have addressed the measurement of distinct dimensions of autonomy, and have confirmed that social institutions of gender powerfully shape women’s autonomy here by region. Findings have not supported, however, the argument that Muslim women are at a disadvantage in terms of women’s autonomy, at least when compared to Hindu women from the same region (Jejeebhoy, 2000; Jejeebhoy and Sathar, 2001). Given that household data from this survey suggest that roughly 75 percent of women were already married by age 18, here we explore marriage patterns among those ages 19 to 39 (39 was the cut-off age for this survey). THE SETTING Uttar Pradesh and Tamil Nadu lie at two extremes of the social and cultural spectrum in India, although economically they are relatively similar. Both states are poor, with about 37 percent in Uttar Pradesh and 40 percent in Tamil Nadu (and 33 percent in India) living below the poverty level, and both states are largely agricultural (Uttar Pradesh, 72 percent; Tamil Nadu, 61 percent; India, 70 percent). Yet there are huge differences in social development levels. For example, literacy rates are much higher in Tamil Nadu (63 percent) than in Uttar Pradesh (42 percent), and fertility and mortality are much lower. The infant mortality rate is 98 per 1,000 live births in Uttar Pradesh and 58 in Tamil Nadu; the under-5 mortality rate is 141 in Uttar Pradesh and 87 in Tamil Nadu; and the total fertility rate is 5.1 in Uttar Pradesh and 2.2 in Tamil Nadu. In most of India, both north and south, and among both Hindus and Muslims, the family is mainly patriarchal, patrilocal, and patrilineal, and the region is well known for inegalitarian gender relations (Altekar, 1962; Karve, 1965). Women are defined as inferior; husbands are assumed to “own” women and to have the right to dominate them. Inegalitarian gender relations deny women a decision-making role in family matters, inhibit
OCR for page 175
The Changing Transitions to Adulthood in Developing Countries: Selected Studies them from moving about freely, prevent their access to material resources, and expose them to violence in the household. Within this situation of generally limited autonomy, however, there are sharp cultural and regional differences in women’s situations and vulnerability (Dyson and Moore, 1983), and these are reflected in available social indicators. For example, in Uttar Pradesh, life expectancy is about 4 years higher for males than for females (54 and 49, respectively); in Tamil Nadu, life expectancy for both females and males is 61 years. Moreover, the maternal mortality ratio is 931 in Uttar Pradesh and 319 in Tamil Nadu. Furthermore, gender disparities in literacy are far wider in Uttar Pradesh (25 percent for females compared to 56 percent for males) than in Tamil Nadu (51 percent for females compared to 74 percent for males). DATA The main objectives of the study from which these data are drawn were to operationalize the concept of autonomy, to assess its relationships to reproductive behavior, and to assess the role of context in conditioning both levels of autonomy and its relationship to reproductive behavior. Samples were drawn from north and south Indian women and from Hindu and Muslim women. The survey included a household questionnaire, a questionnaire for eligible women currently married and ages 39 or younger, and a questionnaire for available husbands. Parallel questions were posed to women and their husbands concerning women’s autonomy within the home, and gender norms and expectations Uttar Pradesh in north India and Tamil Nadu in south India were selected deliberately to represent a range of gender and sociocultural conditions. Within each state, similarly, two districts were purposively selected (on the basis of an index of development, measured from indicators such as income, percentage of roads surfaced, and other economic criteria) to maximize differences in socioeconomic conditions, while at the same time allowing for comparisons of Hindu and Muslim women. From each district, one taluka (subdistrict) was similarly selected. The four sites thus selected were as follows: from Tamil Nadu, Pollachi taluka from Coimbatore district (ranked 1 of 21) and Mudukulathur taluka from Ramnathpuram district (ranked 18 of 21); and from Uttar Pradesh, Kunda taluka from Meerut district (ranked 2 of 63) and Baghpat taluka from Pratapgarh district (ranked 51 of 63). From each of the four sites, a cluster of contiguous villages of roughly 1,000 to 2,000 households was randomly selected, and about 800 currently married women ages 15 to 39 were randomly selected for interview. Husbands who were present were also interviewed. In each setting, on the assumption that sociocultural norms governing female autonomy vary widely among Hindus and Muslims, about half of all respondents selected
OCR for page 176
The Changing Transitions to Adulthood in Developing Countries: Selected Studies were Hindu and the other half Muslim. As a result, a total of eight communities are covered: four geographical sites, and within each site, two distinct religious groups, Hindus and Muslims. A total of 1,842 women constitute the sample: 859 from Uttar Pradesh and 983 from Tamil Nadu. In the course of interviews with women, respondents were asked not only about their education and work status, but also about a variety of dimensions of autonomy within their married lives, including their decision-making authority, personal freedom of movement, control over economic resources, and wife-husband power relations. Also asked were questions relating to marriage age and practices, including the extent to which women had a say in marriage decisions, postmarital residence patterns, the magnitude and content of dowries, and the extent to which women had a say in the use of these valuables. The study was restricted to married women and corresponding data from the unmarried are not available. Ideally speaking, of course, for the study of determinants of age at marriage, unmarried women also must be included in the sample. By performing the analysis on women who are already married by the time of the survey, we allow the sample to be “right-censored”; for example, married women ages 19 to 24 at the time of the survey cannot by definition include women who would be marrying after 24 in the young age cohort, whereas such women would be included in the older age cohorts. The analysis attempts to minimize this limitation as far as possible. This chapter reports findings from 1,715 currently married women, ages between 19 and 39, and married once. A distribution of all females residing in the selected households suggests that by age 19, three-quarters or more reported themselves to be married; hence, we selected age 19 as the lower cutoff point for this study. Given this age range, and because the objective of this chapter is to explore the marriage experiences over time, we compared the experiences of successive cohorts of women by marital duration. While assessing marriage practices, we compared the experiences of the recently married (0 to 10 years) with those who were married 11 or more years prior to the survey. Similarly, in order to assess marital age, we compared the experiences of successive cross-sections of married women as defined by marriage duration. Only once-married women are included. This chapter compares marriage patterns among Hindus and Muslims from each setting, and explores the extent to which marriage age and patterns have changed in the roughly 25 years separating the marriages of the oldest and youngest women. Clearly, the older women will be subject to significant recall lapses that may limit the reliability of their responses. Data are analyzed separately by region (Uttar Pradesh and Tamil Nadu) and religion (Hindu and Muslim) in order to explore their role in influencing differences in marriage age and patterns. Table 6-1 highlights background social and economic characteristics of households of the two cohorts of women by marital duration in the four
OCR for page 177
The Changing Transitions to Adulthood in Developing Countries: Selected Studies communities. Economic status appears to be relatively similar across sites. For example, per capita annual income varies from Rs 2,600 among Tamilian Muslims married 11 or more years prior to the survey to Rs 5,400 among recently married Hindu women in Uttar Pradesh; households, irrespective of cohort or community, possess an average of two to three consumer goods. By and large, land ownership is more likely to characterize Hindu women residing in Uttar Pradesh than other groups, while household amenities, such as electricity, are more likely to characterize the homes of Tamilian women. A look at educational profiles of women and their husbands suggests some interesting trends. Although large proportions of women in all communities have never been to school, and few have completed primary school, considerably larger proportions of Tamilian women have ever attended school than have women in Uttar Pradesh, irrespective of religion. Increases in schooling rates appear to have increased most impressively, however, among Hindu women from Uttar Pradesh. While 11 percent of those married 11 or more years prior to the survey had any secondary schooling, this increased to 41 percent among the recently married cohort. Correspondingly, the husbands of Hindu women from Uttar Pradesh have made the most impressive gains in education—from a mean of 7.4 years of schooling among the cohort married 11 or more years prior to the survey to 10 years among the recently married. Economic activity profiles around the time of marriage differ starkly by region. In general, Hindu women from Tamil Nadu, irrespective of marital duration, are most likely to have worked for wages both before marriage and as newlyweds. Also notable among them is that the more recently married cohort is somewhat more likely than the cohort married 11 or more years prior to the survey to have worked before marriage. In short and in general, this profile suggests some variation in social and economic conditions across the four communities. Within each community, however, marital duration-specific variation is relatively modest, with some notable exceptions, including educational attainment of women and husbands, especially the Hindus of Uttar Pradesh, and wage work before marriage among Hindus of Tamil Nadu. AGE AT MARRIAGE Results presented in Table 6-2 suggest that in general, the familiar north-south dichotomy is reflected in age at marriage patterns; within each region, by religion, differences are modest. Mean and median ages at marriage range between 13 and 17 in Uttar Pradesh and between 17 and 19 in Tamil Nadu. Of interest now are notable changes in marriage age within the four communities studied. Clearly, marriage continues to take place in adolescence for the majority of each cohort under study; however, there is a pronounced shift from early to late adolescence. For example, in Uttar
OCR for page 178
The Changing Transitions to Adulthood in Developing Countries: Selected Studies TABLE 6-1 Profile of Women by Duration of Marriage, Women Married 0-10 and 11 or More Years, Once-Married Women Ages 19-39, Uttar Pradesh and Tamil Nadu Uttar Pradesh Muslim Hindu 0-10 11+ 0-10 11+ Number 149 221 145 259 Land Ownership Status Any land 30.9 33.5 77.2 72.2 6+ acres 2.7 3.6 16.6 13.9 Household Economic Status Mean per capita annual income (Rs) 3,182 2,995 5,380 3,942 % with electricity 37.6 38.9 42.1 25.1 Number of consumer durables owned (of 9) 2.5 2.2 2.4 2.2 Schooling Status 1-6 years 18.8 14.5 20.0 15.8 7+ years 8.7 1.6 40.7 11.2 Mean years 1.6 0.8 4.9 1.7 Schooling of Husband Mean years 5.4 3.8 10.0 7.4 Economic Activity % worked for wages year 1 2.0 0.9 1.4 0.4 % worked for wages before marriage 5.4 2.7 3.4 4.2 NOTE: Excludes a total of 128 women, including 89 ages 18 or less, and 39 older women married more than once. Pradesh, among those married 11 years or longer, two-thirds of Muslims and three quarters of Hindus were married by age 15; this proportion fell to around one-third among the youngest. In Tamil Nadu, where marriage by age 15 was relatively uncommon even among the older cohort (27 percent and 20 percent of Muslim and Hindu women were married by age 15), dramatic changes have continued to occur, with fewer than 10 percent of the younger cohort married by age 15. Marriage by age 18 has also recorded a significant decline in both settings. In Uttar Pradesh, where more than 90 percent of those married 11 years or longer were married by age 18, this proportion has declined to two-thirds among those married more recently. In Tamil Nadu, too, considerable change is apparent: Among Muslims, proportions married by age 18 has declined from 66 percent to 34 percent and among Hindus from 50 percent to 25 percent. Clearly, a powerful north-south divide remains in marriage age and changes in this age. Although marriage in adolescence has shown remarkable declines in both settings, the large majority of women in Uttar Pradesh
OCR for page 179
The Changing Transitions to Adulthood in Developing Countries: Selected Studies Tamil Nadu Muslim Hindu 0-10 11+ 0-10 11+ 230 293 207 211 27.8 33.5 33.3 40.3 2.6 2.4 15.0 11.4 3,292 2,618 4,203 3,587 63.5 64.5 46.9 40.3 3.0 3.0 2.6 2.5 56.5 56.0 40.6 31.3 21.7 7.8 20.8 15.2 4.2 2.8 3.6 2.6 5.7 5.0 4.9 4.6 7.8 12.6 39.1 39.8 21.7 17.4 61.4 55.4 continue to marry in adolescence. In Tamil Nadu in contrast, it appears that marriage is increasingly postponed beyond the adolescent years. MARRIAGE PATTERNS Arranged marriage and extensive dowries continue to characterize marriage in much of India, both north and south. Here we explore the extent to which these traditional patterns are changing over the lives of the two marriage cohorts. Available data enable us to piece together a substantial profile of marriage-related decision making, the extent of arranged marriage and premarital acquaintance between spouses, postmarital residence patterns and extent of endogamy, spousal age and educational disparities, and the size and content of dowry and the extent of female control of its disbursal. Findings suggest that marriage patterns are by and large stable over time, but that practices are far less alienating or likely to limit exercise of autonomy among south Indian women, irrespective of age or religion, than among northern women.
OCR for page 180
The Changing Transitions to Adulthood in Developing Countries: Selected Studies TABLE 6-2 Marriage Age Among Women by Duration of Marriage, Women Married 0-10 and 11 or More Years, Once-Married Women Ages 19-39, Uttar Pradesh and Tamil Nadu Uttar Pradesh Muslim Hindu 0-10 11+ 0-10 11+ Number 149 221 145 259 % married by age 15 34.9 67.4 36.6 76.8 % married by age 18 63.1 93.7 65.5 95.4 Mean age at marriage (women aged 20+) 16.6 14.5 16.7 13.0 Median age at marriage 16 15 17 14 NOTE: Excludes a total of 128 women, including 89 ages 18 or less, and 39 older women married more than once. Marriage-Related Decision Making Marrying off a daughter is regarded as one of the most important duties of a father in both north and south India. Findings presented in Table 6-3 confirm that in neither Uttar Pradesh nor Tamil Nadu are marriage negotiations left entirely to the daughter. Women were asked whether they had selected their own spouses or participated in the decision on when and whom to marry. Responses suggest that for the overwhelming majority, irrespective of marital duration, region, or religion, marriages were arranged either by parents alone or with relatives and matchmakers. Yet subtle differences do emerge, and substantial minorities of women reported having a say, or being consulted in these decisions. While the familiar regional disparities persist, with Tamilian women exerting far more autonomy in these decisions than north Indian women, marital duration-specific differences are also evident within each setting. Recently married cohorts exerted considerably more of a voice than older cohorts in marriage-related decision making. For example, among Hindu and Muslim women from Tamil Nadu, proportions reporting a say in marriage decisions increased from 37 and 29 percent, respectively, among the longer married cohort to 45 and 35 percent, respectively, among the recently married ones. In contrast, among women from Uttar Pradesh, although increases are reported by both Muslim and Hindu women, no more than a handful of even the recently married—about one in six in both groups—had any say or veto powers in this decision.
OCR for page 181
The Changing Transitions to Adulthood in Developing Countries: Selected Studies Tamil Nadu Muslim Hindu 0-10 11+ 0-10 11+ 230 293 207 211 7.0 26.6 4.3 19.9 34.3 66.2 25.1 50.2 18.1 16.8 19.2 17.5 18 17 19 17 Kin Marriage and Premarital Acquaintance with Husband Regional differences in kin marriage and village exogamy have been extensively documented in the literature. In the north, and particularly among Hindus, the practice is to marry young girls into families with which previous contact has been limited and kin marriage is rare—both features that heighten women’s powerlessness (Committee on the Status of Women in India, 1973). In the south, in contrast, both Hindu and Muslim women enjoy less alienating marriage ties. Here, marriage is more a means of consolidating existing kinship networks than a political alliance, and as a result they are both more likely to be married within the wider kin network and to be acquainted with their husbands prior to marriage. Results presented in Table 6-3 bear out these conclusions. Among north Indian Hindus, for example, only 1 to 2 percent were married within the kin network and an equal percentage were acquainted with their husbands prior to marriage. Among south Indians and to a lesser extent north Indian Muslims, kin marriage and premarital acquaintance with husband are far more prevalent. What is particularly interesting from these findings is the suggestion that these are cultural practices that are deeply embedded; there is no convincing evidence, for example, of change over time in either the extent of marriage within kin networks or of premarital acquaintance in all four communities.
OCR for page 182
The Changing Transitions to Adulthood in Developing Countries: Selected Studies TABLE 6-3 Marriage Practices by Duration of Marriage, Women Married 0-10 and 11 or More Years, Once-Married Women Ages 19-39, Uttar Pradesh and Tamil Nadu Uttar Pradesh Muslim Hindu 0-10 11+ 0-10 11+ Number 149 221 145 259 Autonomy in marriage decisions: % had say in marriage decision 14.8 10.0 15.9 6.6 Age difference with husband (mean) 4.6 5.7 4.0 5.0 % husband 10+ years older 5.4 9.5 0.0 3.9 Education difference with husband (mean) 3.8 3.1 5.1 5.6 Husband is 2+ levels better educated than wife 47.7 35.7 57.9 63.3 Husband is as or less educated than wife 42.3 54.3 21.4 25.9 % living in natal family area 16.1 13.6 1.4 1.9 % in polygamous unions 0.7 1.4 1.4 0.4 % married to relative 27.5 23.5 1.4 2.3 % knowing husband before marriage 28.2 19.5 1.4 0.8 % residing with husband’s family after marriage 92.6 83.3 97.9 90.7 NOTE: Excludes a total of 128 women, including 89 ages 18 or less, and 39 older women married more than once. Village Endogamy and Postmarital Residence In the more patriarchal kinship structure prevailing in the north, particularly among Hindus, marriage involves the incorporation of outsiders as wives into the family and the resulting village exogamy ensures a break between the natal family and the family into which a woman is married. In the south, in contrast, marriage is perceived as a means of strengthening existing community-level bonds, and this is ensured through village endogamy (Karve, 1965). Corresponding with findings concerning kin marriage, results suggest wide regional, and to a lesser extent religion-wise, disparities in the practice of village endogamy, measured here by the proportion of women who currently reside in villages in which they were born or spent their childhood. Village endogamy is rarely reported among Hindu women in Uttar Pradesh (1 to 2 percent), is reported by about one-sixth of Muslim women in that setting, and increases to about one-third and one-half among Hindu and Muslim women in Tamil Nadu. Again, marital duration-specific changes are imperceptible, suggesting again that this traditional practice remains the norm. Postmarital residence patterns remain uniform over all communities:
OCR for page 189
The Changing Transitions to Adulthood in Developing Countries: Selected Studies marriage among women. For example, parents’ quest for a mutually acceptable husband is inevitably prolonged when women have a say in whom and when to marry (or at least veto rights) (see, e.g., Dixon, 1975; Mason, 1984). Similarly, where employment opportunities are available for unmarried girls, and work outside the home is socially acceptable and available, women or their parents may defer early marriage in favor of participation in the labor force (Dyson and Moore, 1983; Standing, 1983). Others have argued that the marriage squeeze or relative paucity of males are the factors that have led to an increase both in marital age and education levels of females (Bhat and Halli, 1999; Rao, 1993). In fact, women’s economic independence has been hypothesized as the leading reason for delayed marriage in gender-stratified settings where families stand to gain from the employment of their uneducated daughters, but especially in more equal settings where the woman herself stands to gain (Caldwell, Reddy, and Caldwell, 1983; Jain and Nag, 1986; Mason, 1993; Salaff and Wong, 1977). Others have argued differently, placing education more at the center of the force for change. Factors that enhance age at marriage are said to be related more to reducing the marriageability of females than increasing the autonomy of young women. Even in settings where traditional attitudes and norms persist, the demand for educated wives by “eligible” men has made female education an asset in arranging a good marriage (Minturn, 1984, for Rajasthan; Caldwell, Reddy, and Caldwell, 1982, for Karnataka; Caldwell, Reddy, and Caldwell, 1983, for Karnataka; Vlassoff, 1996, for rural Maharashtra). However, the longer a young girl is kept in school, the longer she will remain out of the net of marriage negotiations. Because the pool of available, better educated potential husbands is smaller, negotiations for the marriage of an educated daughter are, in all likelihood, even more time consuming than for an uneducated one (see Caldwell et al., 1989). This pursuit of an education may deter parents from postponing the search for a husband. To address the limitation associated with the lack of corresponding data on unmarried women, we restrict our multivariate analyses of factors influencing marital age to women ages 19 and above, among whom the overwhelming percentage (more than 95 percent) have made the transition into marriage. To understand the correlates of age at marriage for the two age groups, data were subjected to ordinary least-squares regressions with the age at marriage as the dependent variable. Several dummy variables have been constructed for nominal and ordinal variables; the excluded category is designed to be the reference category by which the effects of other dummy variables can be compared. Table 6-5 presents the results of regression analyses (OLS), which regress age at marriage on various individual-, family-, and community-level
OCR for page 190
The Changing Transitions to Adulthood in Developing Countries: Selected Studies TABLE 6-5 Correlates of Marriage Age Among Women Ages 19-39, Married 0-10 and 11+ Years, Uttar Pradesh and Tamil Nadu: OLS Regression Coefficients, by Site Total Married 0-10 Years Married 11+ Years INDIVIDUAL LEVEL Education Primary 0.986c 1.088c Secondary 2.585c 2.723c Economic activity for wages before marriage −0.222 −0.015 Participation in marriage decisions 0.008 0.218 Marriage Practice and Partner Level Spousal age difference >5 years −0.469b −0.398b Spousal educational difference >5 classes 0.235 0.230 Endogamy: resides in natal village or vicinity −0.099 0.274 Kin marriage: married to relative −0.452b −0.154 Dowry payments >Rs 50,000 −0.008 0.248 NATAL FAMILY LEVEL Father’s Education Level Primary 0.310a 0.138 Secondary −0.048 0.266 Sociocultural Context Uttar Pradesh Hindu −0.892c −2.017c Tamil Nadu Muslim 1.161c 1.481c Tamil Nadu Hindu 2.442c 2.338c Constant 16.244c 14.461c R squared 0.377 0.426 Number 731 984 aEquals .000. bEquals .001-.05. cEquals .051-.099. factors assumed from the above discussion to influence it. Among individual woman-level factors are education, premarital employment, and say in choice of husband. Among marriage practice and partner-level factors are age and educational differences between women and husbands, endogamy, and dowry payments. Natal family factors are represented by father’s education levels, and sociocultural context by state and religion. Results are presented for all women married 0 to 10 years and 11 or more years in Uttar Pradesh and Tamil Nadu, respectively. Findings suggest that not all blocks, and not all indicators within each block, play a role in influencing marriage age. Among individual-level factors, the role of education is prominent. In both settings and in both co-
OCR for page 191
The Changing Transitions to Adulthood in Developing Countries: Selected Studies Uttar Pradesh Tamil Nadu Married 0-10 Years Married 11+ Years Married 0-10 Years Married 11+ Years 1.309b 1.574b 0.992c 0.831c 1.784c 2.812c 3.027c 2.579c −0.027 0.411 −0.074 −0.094 0.337 0.951a −0.159 −0.058 −0.438a −0.017 −0.607c −0.757c −0.207 0.317 0.418a 0.141 0.471 0.006 −0.052 0.100 −0.061 −0.093 −0.517b −0.175 0.068 0.026 −0.163 0.395b 0.090 −0.288 0.436b 0.367b −0.711 0.114 0.747b 0.476a −0.248 −1.962c 1.212c 0.884c 16.308c 14.140c 17.320c 16.270c 0.140 0.150 0.393 0.295 294 480 437 504 horts, even a primary school education has a significant effect on delaying marriage; there is some evidence that in Tamil Nadu, the influence of both a primary and a secondary education on delaying marriage has become stronger among the recently married cohort. At the same time, the effect of a secondary education is clearly far more pronounced in each group. In contrast, economic activity prior to marriage is unrelated to marriage age in each setting and cohort. Thus, having a say in marriage decisions does not appear to delay marriage. Few marriage practice and partner-level factors are uniformly significant. For example, recently married women with considerably better educated husbands are significantly more likely to delay marriage in Tamil
OCR for page 192
The Changing Transitions to Adulthood in Developing Countries: Selected Studies Nadu, but not in Uttar Pradesh. There is no evidence that those marrying within or in the vicinity of the natal village marry any later than other women. In contrast, among both cohorts in Tamil Nadu, and to a lesser extent among the recently married cohort in Uttar Pradesh, findings suggest that women marrying considerably older men (5 years or more) are significantly more likely to marry early than are other women. Among older women—but not younger—in Tamil Nadu (but not Uttar Pradesh), there is evidence that large dowry payments tended to delay marriage. Among natal family factors, there is evidence that natal family socioeconomic status, as measured by father’s education, does influence marriage age, but this effect is statistically significant only in Tamil Nadu. Finally, the evidence suggests that sociocultural context—as measured by religion and setting—exerts a powerful influence on marriage age. A look at findings for the two states combined suggests that compared to the Muslims of Uttar Pradesh, the Hindus of this state marry significantly earlier, but both Muslims and Hindus from Tamil Nadu are significantly more likely to delay marriage. As shown in earlier tables, state rather than religion plays the more prominent role in explaining variation in age at marriage. FACTORS INFLUENCING THE MARRIAGE EXPERIENCE: DOWRY PAYMENTS The literature tends to agree that dowries have become more entrenched in marriage negotiations. One of the arguments posed in the literature is that because there are too few men of marriageable age, families compete for the eligible men by paying higher dowries. A second argument is that a “qualified” man demands a larger dowry or economic contribution from a bride or her family (Jain and Nag, 1986; Kasarda, Billy, and West, 1986), that education has actually strengthened the dowry system (Kapadia, 1958), and that educated women routinely must pay larger dowries than uneducated women (Goldstein, 1970, for Bangalore; Minturn, 1984, for Rajasthan; Seetharamu and Ushadevi, 1985, for Karnataka). In the multivariate analysis of dowry practice, the more conventional logistic regression model is employed, in which the parameter represents the increment or decrement in log odds associated with reports of a large dowry (with rupee values of Rs 50,000 or more), as opposed to small or no dowries (with rupee values of less than 50,000). It is assumed that large dowry payments occur at discrete intervals of time and the odds of conditional probability experiencing the payment of large dowries are small. The analysis is performed separately for women married for 0 to 10 and 11 or more years (ages 19 to 39 and once married). In this way, not only is the
OCR for page 193
The Changing Transitions to Adulthood in Developing Countries: Selected Studies censoring effect reduced, but also we are able to assess changes in covariate effects over two marriage cohorts. Table 6-6 presents the results of logistic regression analyses, which regress payment of a large dowry (equivalent of Rs 50,000 or more) on various factors assumed from the above discussion to influence it. Individual woman-level factors, again, include education, premarital employment, and say in choice of husband. Among marriage practice and partner-level factors are age and educational differences between women and husbands, endogamy, and kin relationship. Natal family factors, as before, are represented by father’s education levels, and sociocultural context by state and religion. Results are now presented for all women by marital duration, on the assumption that marriage practices, if changing, will be reflected among recently married (and not necessarily younger) women. Hence findings are presented for women married 0 to 10 and 11 or more years preceding the survey in Uttar Pradesh and Tamil Nadu, respectively. Findings suggest that individual-level factors play a significant role in influencing dowry payments. As others have noted, dowry payments increase systematically and dramatically by woman’s education level, and especially a secondary school education. What is notable is that the influence of education in one state—Tamil Nadu—has become considerably stronger among women marrying in the recent past than among women who married earlier. For example, compared to uneducated women, seven times as many secondary schooled women who married more than ten years before the survey had paid dowries in excess of Rs 50,000; this ratio increased to 13 times as many among women marrying more recently. A similar trend is observed in Uttar Pradesh, but differences over time and by educational status are more modest. In contrast to the effect of education, premarital economic activity appears to reduce dowry payments for all women, but again, more significantly among Tamilian women, here there is little evidence of change over cohorts. Finally, the results suggest that while participation in marriage-related decisions did not have a significant influence on dowry payments, the direction of the relationship hints that, with the exception of young women in Uttar Pradesh, involvement in marriage-related decisions is consistently inversely related with the magnitude of dowry payments. Marriage practice and partner-level factors also influence dowry payments. For one, findings confirm that a well and better educated husband is more likely to command large dowries than others, although the magnitude of this influence appears to be weakening in both settings. Kin marriages, particularly in Tamil Nadu where they are more prevalent, and notably among the recent cohort, are associated with lower dowry payments. As far as natal family socioeconomic status is concerned, it is clear that father’s education—and particularly a secondary school education—pow-
OCR for page 194
The Changing Transitions to Adulthood in Developing Countries: Selected Studies TABLE 6-6 Results of Logit Analysis of Large Dowry Payments (Rs 50,000 and above) Among Women Ages 19-39, Married 0-10 and 11+ Years, Uttar Pradesh and Tamil Nadu: Odds Ratios Total Married 0-10 Years Married 11+ Years INDIVIDUAL LEVEL Education None 1.000 1.000 Primary 3.536c 2.800c Secondary 8.986c 6.235c Economic Activity for Wages Before Marriage No 1.000 1.000 Yes 0.397c 0.408c Participation in Marriage Decisions No 1.000 1.000 Yes 0.949 0.834 Marriage Practice and Partner Level Spousal age difference 5 years 1.000 1.000 Spousal age difference > 5 years 0.945 1.041 Spousal educational difference 5 classes 1.000 1.000 Spousal educational difference > 5 classes 1.511a 1.971c Exogamy: does not reside in natal village or vicinity 1.000 1.000 Endogamy: resides in natal village or vicinity 1.264 0.578b Not married to relative 1.000 1.000 Kin marriage: married to relative 0.512b 0.668a NATAL FAMILY LEVEL Father’s Education Level None 1.000 1.000 Primary 1.301 1.494b Secondary 3.456c 1.596a Sociocultural Context Uttar Pradesh Muslim 1.000 1.000 Uttar Pradesh Hindu 0.954 1.421 Tamil Nadu Muslim 7.874c 9.711c Tamil Nadu Hindu 4.432c 3.329c Pseudo R Squared 0.26 0.21 Number 731 984 aEquals .000. bEquals .001-.05. cEquals .051-.099.
OCR for page 195
The Changing Transitions to Adulthood in Developing Countries: Selected Studies Uttar Pradesh Tamil Nadu Married 0-10 Years Married 11+ Years Married 0-10 Years Married 11+ Years 1.000 1.000 1.000 1.000 2.515a 1.832c 4.047c 2.370c 5.644c 5.584c 12.669c 7.177c 0.371c 1.000 1.000 1.000 0.738a 0.420 0.376c 1.000 1.000 1.000 1.000 1.029 0.830 0.843 0.812 1.000 1.000 1.000 1.000 1.287 1.065 0.737 1.003 1.000 1.000 1.000 1.000 1.452 2.638b 1.451 1.606a 1.000 1.000 1.000 1.000 1.342 1.589 1.384 0.484b 1.000 1.000 1.000 1.000 2.442a 1.314 0.357c 0.599b 1.000 1.000 1.000 1.000 1.322 3.473c 1.268 0.933 3.307b 2.051b 3.884b 1.156 1.000 1.000 2.005 1.995b 1.000 1.000 0.541b 0.305c 0.21 0.182 0.271 0.192 294 480 437 504
OCR for page 196
The Changing Transitions to Adulthood in Developing Countries: Selected Studies erfully influences dowry payments, and this is even after the influence of the woman’s own education and work status have been controlled. Finally, the influence of sociocultural context—as measured by religion and setting—varies considerably by state. In Uttar Pradesh, the marriages of Hindu women, irrespective of cohort, were considerably more likely to attract a large dowry than were the marriages of Muslim women, although the effect was not significant among the younger cohort. In Tamil Nadu, in contrast, Muslim women are more likely to report large dowries, though there is some indication that differences may be narrowing among the recent cohort. SUMMARY AND CONCLUSIONS The objective of this chapter was to explore marriage patterns and changes in marriage patterns among successive cohorts of women in rural Uttar Pradesh and Tamil Nadu, two culturally distinct social systems. In doing so, it aimed to explore the extent to which marriage age, practices, and factors underlying these had changed in the experiences of successive marriage cohorts. Several conclusions can be drawn from this study, some very clear, and others tentative and suggestive. First, findings confirm that marriage has been increasing moderately and at different paces in the two states. In Uttar Pradesh, among both Muslims and Hindus, early adolescent marriages (under age 15) have declined perceptibly by cohort, yet about two in three recently married women continue to marry by the time they are 18. Among women in Tamil Nadu, in contrast, changes are observed in both early and late adolescence and the evidence suggests that among both Muslims and Hindus, marriage is increasingly being delayed beyond adolescence. This is consistent with other studies (Bhat and Halli, 1999). Second, a review of marriage practices suggests that practices are largely context dependent. Such attributes of marriage as kin marriage, village endogamy and postmarital residence patterns, and spousal age differences, for example, continue to be shaped by region and gender systems. The experiences of recently married cohorts remain largely similar to those of older ones in each setting. However, there are some conditions in which cohort differences have begun to emerge. These include greater autonomy among recently married women in determining the timing of marriage and choice of partner, and greater egalitarianism among these women in terms of educational differences with husband—attributes far more clearly observed in Tamil Nadu than in Uttar Pradesh, and presumably suggestive of more egalitarian relationships among younger cohorts of Tamilian women. On the other hand, a notable change experienced in Uttar Pradesh is in the content of dowry and specifically a greater trend toward consumer goods among younger cohorts. Data provide little clear evidence on trends in
OCR for page 197
The Changing Transitions to Adulthood in Developing Countries: Selected Studies dowry per se—there is a tentative suggestion that it may have increased over time in all four groups, but particularly among the Muslims of Tamil Nadu and the Hindus of Uttar Pradesh. Third, findings confirm considerable similarity in determinants of marriage age in this region, though the magnitude of certain effects may vary. Clearly the most powerful and consistent factors shaping marriage age are woman’s education and sociocultural setting—notably, even the influence of a primary education sharply increases marital age, and the influence of education remains powerful in both cohorts and settings. Sociocultural context—as measured by religion and setting—suggests that compared to the Muslims of Uttar Pradesh, the Hindus of this state marry significantly earlier, while both Muslim and Hindu women from Tamil Nadu are significantly more likely to delay marriage. Also consistent is the suggestion that women marrying considerably older men (5 years or more) are significantly more likely to marry early than are other women; again there is no significant change in pattern over cohorts in this effect. In contrast, few other factors are uniformly significant: Only among the cohort married 11 years or longer—but not the recently married cohort—is there evidence that large dowry payments may have resulted in delaying marriage, and only in Tamil Nadu is there evidence that natal family socioeconomic status has a significant influence on delaying marriage age. Fourth, in contrast, findings suggest considerable heterogeneity in factors influencing dowry payments in the two settings. Findings confirm, for most groups, a strong positive association between payment of large dowries and natal family socioeconomic status, woman’s education, as well as suitability of the husband, measured in terms of his relative superiority in education. The influence of a secondary education appears to have become stronger in the case of recently married women, particularly in Tamil Nadu. In contrast to this effect, premarital economic activity appears to reduce dowry payments for all women irrespective of timing of marriage, but significantly among Tamilian women. Finally, sociocultural context exerts a strong effect on dowry payments. In Uttar Pradesh, Hindus are significantly more likely to report large dowry transfers, and this effect is consistent over cohorts. In Tamil Nadu, in contrast, Muslim women, irrespective of cohort, report significantly larger dowry transfers than do Hindu women. These findings suggest that by and large, sociocultural setting and individual and marriage process-related factors have a powerful effect on determining marriage age and practices. What is not as clear is that these determinants have shifted over successive cohorts of women. Notably, findings suggest that although education plays a significant role in enhancing marriage age, it also tends to raise dowry payments. Conversely, although premarital economic activity is unrelated to marriage age, it plays a significant role in reducing dowry payments. These kinds of findings suggest that strat-
OCR for page 198
The Changing Transitions to Adulthood in Developing Countries: Selected Studies egies to delay marriage and counter the practice of dowry need to expand beyond education or employment. More comprehensive, direct, and context specific strategies must be sought simultaneously—raising community awareness of the negative effects of early marriage and countering fears of allowing girls to remain single; providing for the acquisition of usable vocational and life skills; and enhancing young women’s real access to and control over economic resources and decision making relating to their own lives. REFERENCES Altekar, A.S. (1962). The position of women in Hindu civilization. Delhi, India: Motilal Banarasidas. Amin, S., and Cain, M. (1997). The rise of dowry in Bangladesh. In G.W. Jones, R.M. Douglas, J.C. Caldwell, and R.M. D’Souza (Eds.), The continuing demographic transition (pp. 209-306). Oxford, England: Clarendon Press. Bhat, P.N.M., and Halli, S.S. (1999). Demography of brideprice and dowry: Causes and consequences of the Indian marriage squeeze. Population Studies, 53(2), 129-148. Caldwell, J.C., Reddy, P.H., and Caldwell, P. (1982). The causes of demographic change in rural South India: A micro approach. Population and Development Review, 8(4), 689-727. Caldwell, J.C., Reddy, P.H., and Caldwell, P. (1983). The causes of marriage change in South India. Population Studies, 37(3), 343-361. Caldwell, J.C., Gajanayake, I., Caldwell, B., and Caldwell, P. (1989). Is marriage delay a multiphasic response to pressures for fertility decline? The case of Sri Lanka. Journal of Marriage & the Family, 51(2), 337-351. Committee on the Status of Women in India. (1973). India: Towards equality. New Delhi: Government of India. Das Gupta, M. (1987). Selective discrimination against female children in rural Punjab, India. Population and Development Review, 13(1), 77-100. Dixon, R.B. (1975). Women’s rights and fertility. Reports on Population/Family Planning, 17, 20. Dube, L. (1988). On the construction of gender: Hindu girls in patrilineal India. Economic and Political Weekly, 23(18), WS11-WS19. Dyson, T., and Moore, M. (1983). On kinship structure, female autonomy, and demographic behavior in India. Population and Development Review, 9(1), 35-60. Goldstein, R.L. (1970). Students in saris: College education in the lives of young Indian women. Journal of Asian and African Studies, 5(3), 192-201. International Institute for Population Sciences. (1995). National Family Health Survey: India 1992-93. Bombay, India: Author. International Institute for Population Sciences and ORC Macro. (2000). National Family Health Survey: India 1998-99. Mumbai, India: Author. Jain, A., and Nag, M. (1986). Importance of female primary education for fertility reduction in India. Economic and Political Weekly, 21(36), 1602-1608. Jejeebhoy, S. (2000). Women’s autonomy in rural India: Its dimensions, determinants, and the influence of context. In H. Presser and G. Sen (Eds.), Female empowerment and demographic processes (pp. 204-238). Oxford, England: Clarendon Press. Jejeebhoy, S., and Sathar, Z. (2001). Women’s autonomy in India and Pakistan: The influence of religion and region. Population and Development Review, 27(4), 687-712.
OCR for page 199
The Changing Transitions to Adulthood in Developing Countries: Selected Studies Kapadia, K.M. (1958). Marriage and family in India. Oxford, England: Oxford University Press. Karve, I. (1965). Kinship organization in India. Bombay, India: Asia Publishing House. Kasarda, J.D., Billy, J.O., and West, K. (1986). Status enhancement and fertility: Reproductive responses to social mobility and educational opportunity. New York: Academic Press. Mason, K.O. (1984). The status of women: A review of its relationships to fertility and mortality. New York: The Rockefeller Foundation. Mason, K.O. (1993). The impact of women’s position on demographic change during the course of development: What do we know? In N. Federici, K.O. Mason, and S. Sogner (Eds.), Women’s position and demographic change. Oxford, England: Oxford University Press. Minturn, L. (1984). Changes in the differential treatment of Rajput girls in Khalapur: 1955-1975. Medical Anthropology, 8(2), 127-132. Rao, V. (1993). Dowry “inflation” in rural India: A statistical investigation. Population Studies, 47(2), 283-293. Salaff, J.W., and Wong, A.K. (1977). Chinese women at work: Work commitment and fertility in the Asian setting. In S. Kupinsky (Ed.), The fertility of working women (pp. 81-145). New York: Praeger. Seetharamu, A.S., and Ushadevi, M.D. (1985). Education in rural areas. New Delhi, India: Ashish Publishing House. Srinivas, M.N. (1989). The cohesive role of Sankritization and other essays. Oxford, England: Oxford University Press. Standing, G. (1983). Women’s work activity and fertility. In R.A. Bulatao and R.D. Lee (Eds.), Determinants of fertility in developing countries (pp. 517-546). New York: Academic Press. Vlassoff, C. (1996). Against the odds: The changing impact of education on female autonomy and fertility in an Indian village. In R. Jeffery and A.M. Basu (Eds.), Girls’ schooling, women’s autonomy, and fertility change in South Asia (pp. 218-234). New Delhi, India: Sage Publications.
Representative terms from entire chapter: