ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

The Making of the Middle Class in Western India: Age at Marriage for Brahmin Women (1900-50)

In spite of the recognition of the importance of the middle class, its historical fashioning has not so far been empirically studied. This study is a part of a project that examines the factors that influenced the behaviour of the contemporary middle class in western India using time series data constructed from unpublished sources. A substantially higher age at marriage, which was reached much earlier than other classes, is one of the distinguishing features of the middle class. The current paper examines reasons behind the rapid increase in the marriage age over 1900-50 among the Chitpavan brahmins of Maharashtra. In particular, the project of fashioning the emerging nation, an ideology widely shared, is highlighted as an important factor behind the increase in the age at marriage for middle class women in western India.


The Making of the Middle Class in Western India: Age at Marriage for Brahmin Women (1900-50)

Neeraj Hatekar, Abodh Kumar, Rajani Mathur

In spite of the recognition of the importance of the middle class, its historical fashioning has not so far been empirically studied. This study is a part of a project that examines the factors that influenced the behaviour of the contemporary middle class in western India using time series data constructed from unpublished sources. A substantially higher age at marriage, which was reached much earlier than other classes, is one of the distinguishing features of the middle class. The current paper examines reasons behind the rapid increase in the marriage age over 1900-50 among the Chitpavan brahmins of Maharashtra. In particular, the project of fashioning the emerging nation, an ideology widely shared, is highlighted as an important factor behind the increase in the age at marriage for middle class women in western India.

The authors are grateful to several colleagues and the staff of the Mumbai Marathi Grantha Sangrahalaya, Dadar, Mumbai for excellent support. The usual disclaimer applies.

Neeraj Hatekar ( is with the Centre for Computational Social Sciences, University of Mumbai, Mumbai; Abodh Kumar is with the Symbiosis Centre for Management and Human Resource Development, Pune and Rajani Mathur is with the Department of Economics, SIES College of Arts and Science, Mumbai.

he rise of an ambitious middle class has been commented upon by observers, both academic as well as popular, as a striking feature of contemporary India (Das 2000; Goyal 2003; Rhode 2003). However, detailed empirical case studies of the historical fashioning of the middle classes are not yet common. This study aims to fill a small part of this gap. Indian “middle class” is a rather special case. The middle classes are not necessarily those who are in the middle of the income distribution. In India, the middle classes are defined by their self-image and primarily consist of those upper castes that have been traditionally associated with service castes. Over the last century and a half, the middle classes came to be distinguished by another common characteristic: through following a demographic and family formation trajectory that was different from most other caste groups, but widely shared among the middle classes. The middle classes acquired greater education, married later and shifted to a smaller family norm earlier than other groups. The higher than average acquisition of human capital makes them important drivers of economic growth in the eyes of some analysts (Doepke and Ziliboti 2005 and 2007). However, for India at least, there is a paucity of detailed empirical histories of the making of specific groups into the middle class.


In order to arrive at a detailed understanding of this important group, it is necessary to construct continuous time series on crucial variables and then examine the trends in them in the context of cultural and material developments. In this paper, we have examined the transitions in the age at marriage of one such important group: the Chitpavan brahmins of Maharashtra. This paper argues that the age at marriage for Chitpavan brahmin women, which was an important dimension of distinctiveness, differentiated itself from the rest of the caste groups, and also the Chitpavans of the 19th century, in the decade of 1920-30. We will argue that the ideological and cultural factors were to play a substantial role, over and above the material ones, in getting Chitpavan brahmin women to marry late. Particularly important to this change was the evolving discourse of an emerging nation and the construct of the ideal wife and mother that accompanied the project. This project was particularly at the fore when popular participation in the nationalist movement was at its strongest in the 1925-30 decade, culminating with the Civil Disobedience Movement. This paper argues that the chain of events climaxing in the Civil Disobedience Movement marks a watershed in the history of the age at marriage for Chitpavan brahmins. In this

may 23, 2009 vol xliv no 21

section we examine why we attach particular importance to this event.

The Civil Disobedience Movement (1930-31) began on 6 April 1930 when Gandhi picked up a handful of salt at Dandi in Gujarat to protest the unjust salt-tax. The agitation was soon marked by participation by the masses from all walks of life, including women and children. The call for civil disobedience captured the nation’s imagination like never before. The movement transformed itself into no-tax campaigns, agitation against liquor shops, foreign made clothes and forest laws. The entire country seemed to be steeped in the spirit of civil disobedience (for an evocative account, see Fischer 2007: 331-45).

Women figured significantly; women who had never stepped unescorted out of their homes, women who had stayed in purdah, young mothers, widows and girls, became a familiar sight as they protested firmly but silently outside liquor shops and opium dens or shops selling foreign clothes (Chandra et al 1999: 276). Women were also involved in other ways, in propaganda, illegal meetings and public protests. At least one authority has called the event the most liberating influence on women till date (ibid: 283). It is generally agreed that women’s public participation during the Civil Disobedience Movement had been qualitatively as well as quantitatively different from previous movements like the Non-cooperation Movement.1 Women participated not necessarily overtly; a vast majority of the women who participated did so without actually coming out into the open (Thapar-Börkert 2006: 170-216). The Civil Disobedience Movement ended on 6 March 1931, with the signing of the Gandhi-Irwin pact. Though some contemporaries were disappointed with the ultimate terms of the pact, the Civil Disobedience Movement marks a watershed for popular participation in the national movement.

Mass Nationalist Movements

During the second half of the 1920s several events joined to strengthen the opposition to Britain. The Simon Commission and the death of Lala Lajpat Rai provoked widespread agitation. The arrest of Bhagat Singh and his youthful comrades as well as the death of Jatin Das in jail while on hunger strike could not but arrest national attention. The acceptance of the demand for complete independence by the Congress and the election of the charismatic Jawaharlal Nehru to its presidency further underlined the opposition to imperial rule. The Congress and Gandhi personally toured the length and breadth of the country in order to prepare it for civil disobedience (Chandra et al 1999: 264-65). On 26 January 1930, public meetings were organised all over the country where the independence pledge, listing India’s grievances against England, and pledging complete freedom from British rule as the goal, was read out in the regional language. The national flag was hoisted following the public reading of this pledge. This train of events, as well as the degree of preparation, separated the Civil Disobedience Movement from the Non-cooperation Movement. Also, the Civil Disobedience Movement came at the time of the Great Depression when agricultural prices had plummeted. Prosperous landed interests joined the movement demanding non-payment of agricultural taxes (Misra 2007: 191). The Civil Disobedience Movement was not the first mass

Economic & Political Weekly

may 23, 2009 vol xliv no 21

m ovement during the Indian nationalist struggle. The Noncooperation Movement of 1920-22, perhaps, deserves this label more. However, the Civil Disobedience Movement was more significant both in terms of the actual numbers that participated as well as the wider ideological impact on the public imagination. The statistics speak for the difference. During the nearly two years of the Non-cooperation Movement, 20,000 were arrested. In less than a year of the Civil Disobedience Movement, the number of arrests was nearly 60,000. The participation was so wide that newspapers announced the list of essentials that volunteers should carry to avoid inconvenience during their incarceration. The Civil Disobedience Movement therefore was a much bigger event and marked the climax of the most significant phase of popular participation in the nationalist movement.

Despite the recognition of women’s mass participation in the freedom movement, the connection between the way women were constructed in the public sphere and concrete demographic changes like changing age at marriage has not been explored so far. The public construction of “ideal wives and mothers” was closely tied to the emerging construct of the nation. This paper focuses on changes in the age at marriage. Changes in the age at marriage are significant for a number of reasons. For one, a substantial amount of historiography has focused on the evolution of legislation governing marital relations. Laws governing age of consent and age at marriage have been the focus of much social history.2 However, whether the actual age at marriage was influenced by the widely shared ideas about ideal wives and mothers has not been explored.

This paper is divided into four sections. Section 1 is the Introduction. Section 2 explores the link between the conceptions of ideal wives and mothers and the public construction of the appropriate age at marriage. We argue in Section 2 that the evolving image of the mother and wife in the nationalist discourse implied a gradual increase in the age at marriage. The Civil Disobedience Movement as a climactic moment of popular nationalism, underlined the nationalist construct very strongly. As a result, we may expect the actual age at marriage to show some impact of the Civil Disobedience Movement. This is investigated in Section 3 where we present our data and the econometric results. The data consist of a continuous time series created from family genealogies in western India from 1900 to 1947. After controlling for a number of other factors that can be expected to impact the age at marriage, we find that a significant increase in the age at marriage for women is associated with the Civil Disobedience Movement. The last section concludes the paper.


In western India, at least from the mid-19th century, national “upliftment” was being couched in terms of the reforms of the household and conjugal relationships. Early marriages had drawn particular attention as an institution in urgent need of reform. Direct links were drawn between early marriages and national well-being. The author of the prise winning essay submitted to Bombay University on the subject of early marriages had brought out the supposed connection between early marriages and poor health as early as 1876 (University of Bombay 1876: 45). In the


1880s, a social reformer from Bombay, B M Malabari, proposed to the government, several measures designed to curb child marriages. The government promptly invited comments from leading members of the public. M Tillainayagam responded, “Besides, early marriage affects the general health of the married couple and their progeny, and the result is we have a weak and imbecile nation” (Government of India 1886: 28). Other effects attributed to child marriage included retardation of intellectual progress, sapping the foundations of mental and moral energies, checking of the spirit of travel in foreign lands, ill-matched alliances, the evil of overpopulation and enforced widowhood (ibid: 69).

Motherhood and Sound Progeny

By the 1930s, the connection between enlightened motherhood and national rejuvenation had become quite routine. The idea of an emerging nation, and the need to have mentally and physically sound progeny, seems to have driven the demand for delayed marriages. In 1910, the family magazine Manoranjan invited its readers’ opinion about the age at which it would be appropriate to marry. Out of 24 replies 23 were in favour of delayed marriages, the most common reason cited being the potential harm caused to the nation by the offspring of mentally and physically underprepared parents.

Stories emphasising the relationship between reforms of domestic relationships and national revitalisation were appearing in magazines read by middle class men and increasingly, by women. For example, Manoranjan published a story titled “Grihini Sachivah” in December 1929 (Kulkarni 1929). In this story, the husband is shown living in a happy marital paradise with an educated wife who is also an excellent cook and who enjoys literature and the theatre. He seeks her opinion about literary and artistic matters and feels enlightened. But he keeps her away from financial matters. The wife persuades him, very gently, to allow her to look after the household finances too, and finally rescues the family from a long-cheating moneylender. The story ends with the plea that all husbands ought to allow their wives to handle domestic finances. Such stories were commonplace at the time.3 It is interesting to note that the discussion about handing over financial control to the wife is couched in the language of the imperial government gradually expanding political autonomy to Indian men. National politics and the domestic sphere seem to reflect one another. These women are critically placed such that the responsibility for reforms in the family, and thereby the nation, rested with them. As the Education Report 1922-27 states,

The women of India of all communities are rapidly awakening to the urgent necessity of educating their daughters not necessarily for employment or high scholarship but at least to be able to take a more intelligent share as mothers and wives in the training and upbringing of children and in the daily affairs of rural and urban life (Government of India 1929a: 158).

Women were now being imbued with agency that was vital, yet in keeping with their “feminine” positions. In several essays by women writers of the 1920s and the 1930s, women hold the primary responsibility for suprajajanan, giving birth and molding the “ideal” citizens (Sukathankar 1935: 257; Krishnakumari 1927; Kende 1927). The February and March 1919 issues of Manoranjan carried two articles, both by highly educated unmarried women. The first was written by Indira Kelkar, BA, while the second was authored by Durgabai Nene. The articles reflect the idea that an “ideal” nationhood could only be based on an “ideal” family. Indira outlines a deeply moral project of nationhood, the cornerstone of which is the woman, both in her role as a mother inculcating the appropriate values in children, as well as a wife who is “well matched” with her husband, in the sense of being able to fully participate in the nation-building programme. This, according to Indira Kelkar, was precluded if the wives were too young (Kelkar 1919). The Joshi Committee, appointed to examine issues related to the age of consent, also thought that the intellectual compatibility of the wife with the modern husband was impossible in early marriages (GoI 1929b: 169).

A reconstruction of the physiology of women was emerging in order to suit the project of women as sites where the new nation was to be reproduced. D K Karve, the founder of the first women’s university in India, was particularly impressed by the following paragraph from a Japanese information brochure:

In a recent lecture to the Ethnological society, Dr Bernard Hollander gave a very clear explanation of some of the main characteristics of the typical woman, and related these to the specific structure of the body, brain and nervous system. Owing to greater relative importance of the sympathetic nervous system in women, and its greater fineness, women are more capable than men of experiencing and expressing feelings such as joy, grief, hope and are more instinctive and subjective in their reactions. Men will never equal women in intuition, quick receptivity, adaptability and “emotionality”. On the other hand, the greater stability of the male nervous system is the result of a different constitution. Nature knows her own job, and if women were in nervous constitution like men they would be unfitted for their racial function (Karve 1929: 482).

An article in the special issue of a popular middle class magazine dedicated to women’s education emphasised how forcing girls to study subjects meant for boys compromised their primary function of reproduction by diverting blood supply from the reproductive organs to the brain (Chiplunkar 1929). Women were constitutionally suited only to become homemakers. Their s pecific role in the nation-building process was to act as sites of reproduction of the next generation of men inculcated with n ationalist values (Karve 1929: 489).

Ideal Mothers and Marriage Age

Women, in order to be able to imbibe the requisite skills to be ideal mothers and wives, had to be of a certain age before they married. Opinion was changing in favour of a higher age at m arriage. Towards the end of the 19th century, most brahmin households thought it a supreme religious obligation to marry off their daughters before they reached puberty. The scene had changed by the 1930s. The extent of the change is obvious when one compares the tone of discussions in the 1920s with the respon ses received to the Malabari proposals. As K Krishnaswamy Rao, chief justice of Travancore had pointed out in his reply to the Malabari proposals in 1884, “I may, without fear of contradiction, add that the orthodox element still represents not less than 99% of the native communities; that nearly 90% of those who are convinced of the reasonableness of, or the necessity for, the reforms,

may 23, 2009 vol xliv no 21

Figure 1: Estimated Age at Marriage (Men and Women) (1900-47)

35 30 25 20 15 10 5 0

Estimated mean age for men Estimated mean age for women

Year 1906 1912 1918 1924 1930 1936 1942 1948

are not likely to act up to their convictions for several decades to come” (GoI 1886: 11). Almost everyone, even die hard reformers whose opinions had been sought over the Malabari proposals, had feared that legislation would create resistance; spread of education, particularly women’s education seemed the sure way forward. On the other hand, by 1929, the Age of Consent Committee (1929) writes in its report,

It is because witnesses have realised that the advance by education and social reforms will be so slow and will take such unreasonably long time that they have advised the committee to recommend a more expeditious and effective method of achieving the end, namely, legislation (GoI 1929b: 153).

The urgency seems to have been prompted by the special skills that the new age mother and wife was required to have for building the new nation. As the committee further put it:

She must have enough knowledge to conduct a home and to bring forth and rear up children and even educate them to some extent. She may have preferably more knowledge about her duties to society, but in any case, she must have this minimum. A certain period is necessary for the purpose and sixteen years of the life of an ordinary girl would be fully taken up to acquire this minimum knowledge (ibid: 170).

The potential mother had also to be old enough so that “further physical deterioration of the people can be prevented” (ibid: 216). The June 1929 issue of the Kirloskar magazine carried the results of the competition for the best essay on the “Ideal Family”. The “ideal family” cup was won by one Yashodabai Bapat from Yawatmal. The ideal wife of the essay

Figure 2: Secondary School for Girls (1900-47)

wins her husband’s heart by thrift, or 200
derliness, and by learning English, cy 150
cling and harmonium in her spare time, as well as by participating in the wom 100
en’s club (Bapat 1929). These attributes 50
were being reflected in the list of char 0 1901 1907 1913 1919
acteristics that a “desirable” bride to be

had to exhibit. The matrimonial advertisements of the time have girls to be wed who were not only educated, but could stitch, embroider and play the harmonium as well.4 A set of matrimonial advertisements for 1929 has the average age of a girl with all such desirable qualifications already at 17-18 years, with some as old as 21, while the youngest was 15. This would have been inconceivable three decades ago.

Partha Chatterjee (1999) argues that in response to colonial criticism of the treatment of women in India, the nationalists embarked on a programme of cultural emancipation within the domestic sphere, which, in turn, was carefully shielded from

Economic & Political Weekly EPW may 23, 2009 vol xliv no 21

negotiations with the colonial government. This project required women with new social forms of disciplining – orderliness, thrift, personal sense of responsibility, practical skills of literacy, accounting, hygiene, an awareness of the outside world, characteristics that were clearly incompatible with very young brides. Others argue that reformers sought to redo families along the lines of what they perceived as ideal high Hindu forms (Devika and Mukherjee 2007: 102-28). Whatever may be the motivation, the rhetoric of the emerging nation and women’s supposed role in building that nation was contributing to a desire for later marriages.

From the foregoing, it is clear that at least at the level of ideology, one can posit a connection between nationalistic ideals of women and the age at marriage. There is a substantial body of literature analysing nationalist construction of women at a cultural-ideological level.5 Was the ideological appeal strong and pervasive enough to have had a statistically significant influence on the actual age at marriage? The Civil Disobedience Movement was the striking climax of the series of events in the 1920s that provided a mass base to the nationalist movement. It marked the culmination of a significant phase of popular participation in a myriad ways. It might be worth examining the relationship between this event and the age at marriage for women. Is there evidence of a significant upward shift in the age at marriage that is associated with the Civil Disobedience Movement? If the answer is no, it would imply that, in spite, of a public avowal of the need for later marriages, households, in their private spheres, continued to stick to traditional practices of early marriages and that the public ideological stance regarding women had limited permeability into the household’s private domain. On the other hand, if the answer is yes, we would be uncovering an important relationship in Indian social history. We would be demonstrating an important way in which the Indian national movement impacted the lives of common men and women.


In this section, we set out to empirically examine whether the Civil Disobedience Movement had a significant impact on the

actual age at which women married. Our data consist of family genealogies of Chitpavan brahmin households from Maharashtra, the socalled kulvritantas. This fairly exhaustive data source has not received

1925 1931 1937 1943 the attention that it deserves, with

only one historian taking it seriously so far.6 Since the second decade of the 20th century, Chitpavan brahmin clans have been publishing their genealogical records. The Chitpavans migrated all over India during the 18th and the 19th centuries. In the 18th century, they migrated out of Konkan, the narrow coastal strip in western Maharashtra to spread all over the Maratha dominions as priests, soldiers and administrators. After the onset of the British rule, they migrated as clerks, lawyers, doctors, and generally white-collared employees all over the country. The basic impetus to writing family genealogies was to put together information that would inform the different branches about each other. An added objective was to inform


them of the “glorious deeds” of their ancestors. Till date, several dozen kulvritantas have been written and more are in process. They are available in printed form and are easily accessed though most were published in small editions and circulated privately. Overwhelmingly, they concentrate on

Table 1: Five-Year Averages of Growth Rates of Age at Marriage (in %)

Year Average Growth Rate for Women Average Growth Rate for Men

1900-05 4.06 5.69

1906-10 1.14 -1.78

1911-15 1.40 1.43

1916-20 0.68 3.55

1921-25 0.23 -0.29

1926-30 4.63 1.91

1931-35 1.56 0.78

Equally importantly, the concept of sanskritisation describes “the process by which, a low caste gives up its own rites, and beliefs, and takes up, instead, the customs, rites and of a higher caste” (Srinivas 1956: 79). Lower caste groups increasingly adopt the cultural practices of brahmins, including marital

the male lineage. We could find only 1936-40 1.14 1.49 ones. Sanskritisation has been extre
17 that had information on women’s 1941-45 1.53 3.13 mely productive in explaining the
birth and marriage years.7 Typically, 1946-50 0.90 -2.95 adoption of brahminical practices by
the following information on any given 1900-50 1.66 1.19 other caste groups. The cultural prac

woman is to be found: (a) the year of birth, (b) the marriage year, (c) level of education if any, (d) the dates of birth of male and female children, (e) the birth date of the husband, and (f) father’s and husband’s occupations and addresses. In this paper, we have used this information to compute the age at marriage for men and women for the period 1900-47.

Caste Hierarchy in Sample

It is necessary to underline that our sample is restricted to a caste that is at the top of the caste hierarchy and which constitutes a minuscule fraction of the total population.8 It is important to keep the salient features of our sample in mind. The households considered in this sample have a much larger proportion of urban households compared to the all India average. The literacy level too is much higher than average. There are clear economic status differences too. In the sub-sample of rural households, agriculturists are typically landowners with no landless workers. The other rural occupations of men are mostly restricted to teaching, priesthood and shopkeeping. Broadly speaking, 55.5% of the males in the kulvritantas are employed either in government or private service, while nearly 22% have independent occupations, comprising mainly of small shops and independent businesses including medical and legal practice. There are 16% agricultu rists, while another 5.5% are in religious occupations. About 1% do not report their occupations (Patterson 1970: 385). This occupational structure is quite distinct from the rest of the country where even by the end of our study period, more than three-fourths of the population was involved with agriculture. It is obvious that our sample is not representative of the all India population.

The fact that our sample consists only of women belonging to a specific caste by itself is less damaging than appears at first. For one, the ideological influence of the Chitpavans is quite out of proportion to their numerical minority. For most of the period under review, they exercised a disproportionate share of influence in politics, education, society, social reform, journalism and literature in large sections of western India and in persons like Mahadev Govind Ranade, Gopal Krishna Gokhale and Bal Gangadhar Tilak, achieving recognition throughout the country. The Chitpavans also held a disproportionate share of government posts as administrators (Johnson 1970: 105-07). Since brahmins dominated the publishing industry (accounting for nearly two-thirds of all editors between 1901 and 1921), their cultural ideals had a wide influence. Almost the whole of the literature referred to in section one involves constructs of women who are situated in this group.

tices of the Chitpavans were seen as models to be imitated by other groups. In that sense, understanding what was happening to the age at marriage of Chitpavan brahmin women is not as restrictive as may seem at first.

We have collected data on 1,622 women and 1,282 men married between 1900 and 1947. Instead of estimating the age at marriage indirectly as the singulate mean age at marriage a la Hajnal (1953) our data allows us to estimate the age at marriage for a given year directly, as long as we are willing to parameterise the distribution. We assume that the age at marriage in any given year follows a Weibull distribution with possibly changing parameters for different years. We have estimated the mean age at marriage as the maximum likelihood estimate of the mean of the Weibull distribution for each year.9 Given that for most years, our sample is not very large, a small sample correction has been used (Hirose 1999). Figure 1 (p 43) shows the estimated age at marriage for men and women during our period.

The age at marriage for girls increased from 10.3 years at the beginning of the century to 21.4 years at the time of independence. This is much faster than the rise in the mean age at marriage for the country as a whole. The mean age at marriage for Indian girls was 13 years for 1901-11 (Guttmacher Institute 1987). The average mean age at marriage for girls in our sample for this period is 12.44 years. By 1947, the average age at marriage for girls in our sample had risen to 20.4 years. On the other hand, even by 2001, the average age at marriage for Indian girls as a whole was 18.3 years (Census 2001). For Maharashtra taken as a whole, the average age at marriage in 2001 was 18.8 years for females, which was reached by girls in our sample 65 years earlier in 1936. Our sample shows a doubling of the age at marriage over five decades. Brahmin girls started the century with an age at marriage very close to the national average, but outstripped the rest very rapidly during the next five decades. For Indian men, the mean age at first marriage for 1901-11 was 20.3 years. The mean age for 1901-11 for our sample is 20.5 years. However, by 1947, the sample mean age at marriage reaches 28.5. By contrast, till 2001, Indian men, on an average, had not yet reached 24 years as the age at first marriage.

A closer examination of the data brings out the importance of the year of civil disobedience. Table 1 shows the five-year average growth rates of the age at marriage for women as well as men from 1900 to 1945. The fastest growth for women took place in 1925-30, when the age at marriage grew at 4.63% on an average, as compared to the period average of 1.66. Indeed, after 1925-30,

may 23, 2009 vol xliv no 21

the average growth rate for the age at marriage for women is closer to its period average. As we have pointed out in Section 2, 1925-30 was also the period when a substantial mass base was created for the nationalist movement. In particular, the average age at marriage jumped by nearly two years, from 15.99 in 1929 to 17.97 in 1930. This is the second single largest absolute increase of any year in our data. The largest jump in the age at marriage for women took place in 1943 when the age at marriage jumped from 20.04 years to 23.32 years. This was almost certainly due to the opening up of various war time employment opportunities for women and fits well with the qualitative evidence (Bedekar 1947: 359). However, the 1943 increase proved to be rather shortlived. After one year, the age at marriage was back to 20.4 years and never exceeded 20.6 till the end of the study period. On the other hand, the increase in 1930 was more durable. During the decades of the 1930s, there was a nearly continuous increase in the age at marriage till it crossed 20 years in 1941.

Impact of Education

What were the factors behind the increase in the age at marriage in our sample? To understand the relationship between ideological factors and the age at marriage, we must look at other nonideological factors. One possible explanation that needs to be examined is whether the age at marriage was rising because job opportunities became available to women, which required them to accumulate increasing amounts of education before marriage. It is true that the number of brahmin girls attending educational institutions was increasing during this period. The number of brahmin girls attending colleges in Bombay presidency increased from 106 in 1925 to 201 in 1932. Similarly, the number of girls in secondary schools increased from 1,172 to 3,256 while those in primary schools increased from 28,464 to 41,795 in the same p eriod (GoI 1932: 250; GoI 1925). But the numbers in professional education continued to be minuscule

Figure 3: Estimated Marriage Squeeze (Three Year Moving Average)

in proportional terms. The number

of women studying to become law 1.5
yers was just one in 1925, which in 1
creased to three in 1932. The number
of women studying medicine in 0.5
creased from six to 27 in the same 0
period. In terms of professional edu 1901 1908 1915 1922
cation, the number of girls had not

become large enough to substantially impact the average age at marriage. Labour market returns driving the age at marriage upwards does not look like a plausible hypothesis on this cursory examination of the data. Women essayists writing on the need to educate women also seem to be fairly clear in their minds about the ultimate aim: to help women fit better in their role as mothers and wives rather than take up jobs (Bole 1921; Bedekar 1976: 18). In contrast to married women, widows looked at occupational training as a way to financial independence. In women’s educational institutions, majority of those who acquired training in nursing or tailoring after completing seven years of schooling were widows (Bedekar 1976: 17-18). For other women, education was not to lead to a job. In fact, women working out of their homes would be stigmatised.

Economic & Political Weekly EPW may 23, 2009 vol xliv no 21

Another determinant of higher educational attainments and hence later age at marriage could have been physical access to schools. As schools become more accessible, the cost of schooling goes down, leading to girls spending more time in them and out of the marriage market. There were no special schools for Chitpavan brahmin girls. Here we will focus on the availability of secondary schools for girls in general (even though co-education schools were important in Bombay presidency), because as the age at marriage had already crossed 10 in the beginning of the 20th century, access to primary schools would be irrelevant to increasing the age at marriage. The number of secondary schools for girls was rising in this period. There were only 65 secondary schools for girls in 1901, which increased to more than 170 at the time of independence. Figure 2 (p 43) shows the number of secon dary schools for girls in the Bombay presidency over the study period.

The expansion in the number of schools, especially since the late 1920s, is evident from Figure 2. This might have had an effect on the age at marriage.

The Marriage Squeeze

Demographic literature has emphasised the idea of “marriage squeeze” as an important determinant of the age at marriage (Glick et al 1967; Caldwell et al 1983; Bhat and Halli 1999). Marriage squeeze implies the inability of the marriage market to clear because of an imbalance between the number of prospective grooms and brides. For societies where men marry at a later age than women, demographic expansion creates a surplus of brides at the marriageable age for girls. This leads to some brides postponing the age at which they marry simply because they cannot find suitable grooms. This phenomenon has been studied in the Indian context using census data, and a relative shortage of males has been documented for the period of our study. The existing studies have used the decennial census data to estimate marriage squeeze. A

1929 1936 1943

useful measure of the marriage squeeze is given by the modified weighted sex ratio. This measure is a ratio of the weighted mean of the age distribution of unmarried males to the weighted mean of the ages of unmarried females. The weights used are the age-specific marriage rates (Verma 2003: 183-84). We have modified this measure to suit our data as follows:

Let τ be the age of the youngest male marrying in year t.


For each of the males marrying later than τ in year t, we have MF = ∑i where Niis the number of males marrying at

t≥τ Pit *Nitt age i in year t and Pi is the associated probability for first marriage at age i in year t. A similar measure is obtained for females, FFt. The probabilities for first marriage at all i are obtained by using non-parametric Kaplan Meier estimates of hazard functions for each of the years (see Cameron and Trivedi 2005: 581 and Appendix 1). The index for marriage squeeze for the year t is obtained as MF/FF. These estimates are further adjusted for


differential reporting incidence of male and female marriages by multiplying by the ratio of the number of females to males that were reported in year t in the sample. Figure 3 shows the three-year centred moving average of this estimate.


The estimated marriage squeeze points to persistent shortage of grooms through our period. Figure 3 indicates that this s hortage was gradually easing out till the end of the 1930s, after which it worsened again. The easing of the marriage squeeze in our case is due to men moving into high marriage probability ages relatively faster than females.

Another factor delaying marriage for women would also be the increasing age of the grooms. It is not unreasonable to assume that older and more educated grooms would be looking for more “companionate” and mature wives. Hence, we argue that the age at marriage for men should also be an important determinant of women’s age at marriage.

Approximately 50% of the men and women in our sample belong to households that are in the services sector. Consequently, the share of service sector in gross domestic product (GDP) is used as an indirect measure of the economic status of these groups. An increase in this share is taken to stand for a larger set of economic opportunities, which might affect the age at marriage.

The data for all the above variables were all logged and tested for a unit root using the Phillips-Perron test at 5% level.10 It was found that the null hypothesis of a unit root could be rejected for all the variables except the number of schools. The null of a unit root could be rejected for the first difference of schools. We are now in a position to set up an econometric model: LAGEWt = α1 + α2 * LAGEM + α3 * LMVSQZt + α4 * DLSCHLSt


+ α5* LTERS + α6 * DUMMY+ et

tWhere LAGEWt is the natural logarithm of the estimated age at marriage for women in year t, LAGEMis the natural logarithm of


the estimated age at marriage for men in year t, LMVSQz is the natural logarithm of the three-year centred moving average for the estimated marriage squeeze; DLSCHLS is the first difference


of the natural logarithm of the number of schools in the Bombay presidency for year t, LTERS is the natural logarithm of the share


of the tertiary sector in GDP (at constant prices) for year t, and DUMMY is the all important variable which measures the impact of the Civil Disobedience Movement. This variable takes values 0 from the beginning of the sample period to 1929 and 1 from 1930 to the end of the sample period. We present the estimated model in Specification 1.

Adjusted R-square = 0.90. There was no serial correlation among the residuals

Specification 1

as indicated by the Variable Name Estimated Coeff t-value p-value

Jarque-Bera test for LAGEMt 0.371 2.72 0.010
up to 12 lags. Because LMVSQZt 0.098 2.7 0.010
of the obvious pro- DLSCHLSt -0.026 -2.84 0.007
blem of heteroskedas- LTERSt 0.651 5.70 0.00

DUMMY 0.106 3.3 0.002

ticity here, the results

Constant -0.717 -1.32 0.191

use White’s heteroskedasticity consistent variance-covariance matrices.

All variables in the model are significant at 5% and have the theoretically correct expected signs. The sign on the marriage squeeze is positive, indicating that a relative abundance of grooms was associated with increased age at marriage for women, contrary to the general argument in the marriage squeeze literature. In that literature, a shortage of grooms leads to some women postponing marriage. However, such an interpretation neglects the cultural context of the marriage market. We have seen that in our period, the marriage squeeze was easing out because more and more men were moving into the high marriage probability ages at a rate that was relatively faster than the rate for women. These older men naturally looked for more “modern” wives, who by the argument of Section 2, had to be older than in the past. The positive sign can be explained by assortative mating. The particularly important finding of the regression is that the dummy for the Civil Disobedience Movement was significant and positive, lending support to our hypothesis.

As a way of testing the robustness of our results, we modify our measure of the age at marriage. Instead of estimating the average age at marriage using the Weibull distribution, we use the simple arithmetic mean of the observed ages at marriage for each of the years in our sample for men as well as women. The correctness of the age at marriage estimates depend on the extent to which we have correctly identified the underlying likelihood function. All the individuals in our data are ultimately married. Hence, the shape of the Kaplan-Meier non-parametric estimates could be largely a construct of our data. This can lead to misleading results. The use of the arithmetic mean as a measure for the age at marriage would allow us to check the robustness of our results to an alternative specification. The specification of the r egression equation is the same as Specification 1, except that the measure of the age at marriage is the arithmetic mean of the sample observations. The results are shown in Specification 2.

The adjusted R-

Specification 2

square was 0.93. The

Variable Name Estimated Coeff t-value p-value

regression was cor-

LAGEMt 0.48 3.98 0.000

rected for autocorre-

LMVSQZt 0.08 1.94 0.059

lation. The dummy DLSCHLSt 0.005 0.42 0.67
variable comes out to LTERSt 0.56 4.45 0.000
be significant though DUMMY 0.08 2.53 0.025
its coefficient is much Constant -0.80 -1.718 0.094

smaller now.

The second world war created several job opportunities for women working as clerks and similar white-collared jobs. This could have impacted the age at marriage. Hence, we reran the two specifications with a dummy variable which equalled zero from the beginning of the sample period till 1940 and one thereafter to account for this effect (DUM 43). All the specifications were rerun with the growth rate of the tertiary sector (TERTG), the logarithm of the share of the secondary sector (LSECS) and its growth rate (SECG) as well included in the equation. These v ariables were included so as to control for the general transformation of the economy over the years. The results are given in Appendix 1 as Tables 1 and 2. The civil disobedience dummy continues to be significant. In Appendix 1 Table 2 it is significant at slightly over 7%, while in Appendix 1 Table 1 it is significant at less than 5%. The coefficient of the dummy variable also does not swing wildly. We find that the dummy variable is quite robust to alternative measurements of the age at marriage and alternative specifications.

may 23, 2009 vol xliv no 21

Finally, we checked both the specifica-Appendix Table 1: Dependent Variable – Log of Restraint Act (1929) are highly pessimistic. Women’s Age at Marriage (Weibull )

tions for a structural break in the equation Some even argue that the age at marriage

Adjusted R -square=0.91

in the sense of a change in the relationships actually went up in the six months before

Variable Coefficient p-value

between the dependent variables and other the passing of the Act (Whitehead 1995).

LAGEMt 0.2054 0.079 regressors excluding the dummy. If such a LMVSQZt 0.2750 0.002 Indeed, there were s everal public violations

change has taken place post-1930s and has DLSCHLSt -0.033 0.001 in protest against the Act which went un

been neglected in our analysis, our regres-LTERSt 0.71301 0.00 punished ( Kesari, 15 April 1930).

sion will be mis-specified. Consequently, we reran both the specifications with slope dummies post-1930 for all the regressors, but without including the civil disobedience dummy. We then applied a standard F test of zero restrictions on the slope dummies post-1930. The null h ypothesis of no structural change could not be rejected in both the specifications at 5% level of

DUMMY 0.094 0.031
DUM43 0.072 0.094
LSECSt -0.1917 0.051
SECGt 0.0019 0.149
TERTGt -0.0034 0.052
Constant -0.1129 0.858

Appendix Table 2: Dependent Variable – Log of Women’s Age at Marriage (Arithmetic Mean )

Adjusted R-square=0.93
Variable Coefficient p-value
LAGEMt 0.3449 0.010
LMVSQZt 0.21123 0.022
DLSCHLSt 0.0037 0.757
LTERSt 0.6787 0.00
DUMMY 0.072 0.072
DUM43 0.033 0.248
LSECSt -0.1038 0.281
SECGt 0.0014 0.247
TERTGt -0.0036 0.035
Constant -0.66 0.209

On the other hand, some contemporary administrators found the law to be effective. For example, on being asked whether the law had deterred child marriages, the collector of Thar and Parkar in Bombay presidency replied “The number of child marriages within the prohibited age-limits has very considerably decreased and is on the wane, though it cannot be affirmed that they are entirely extinct” (Baines 1931: 176). The c ollector of East Khandesh replied “My own view is that there has been a big decrease in such marriages and the Act has been a great social advantage” (ibid). But not all were so optimistic. The collector of Surat prophesied “The law will either become a dead letter or will dwindle down to a belief that young children can be married by those who are prepared to pay the fine” (ibid: 175). The


The Sarda Act

We have fairly robust evidence to argue that something happened in 1930 which had a significant impact on the age at marriage. However, 1930 saw another factor that could have impacted the age at marriage permanently – the Age at Marriage Restraint Act (1929) became effective from 1 April 1930.

The final enactment of the Sarda bill raised the legal age at marriage for girls to 14. This bill came to be known as the Sarda Act. It provided for a fine of Rs 1,000 in addition to imprisonment up to one month for adults solemnising the marriage of a girl under 14 years of age. Groom over 21 years of age could be imprisoned. In spite of these provisions, the credibility of the penal clause was doubtful. The fine could hardly be a deterrent if it was considered as just another expense connected with the marriage. It is unclear who would have complained against such a marriage. Complaints lodged within one year of the marriage alone were valid. The complainant had to execute a bond of Rs 100 to compensate in case the complaint turned out to be false. Women, even if convicted, were not to be imprisoned. Adults solemnising such a marriage could always argue that they were “under the impression” that the girl was over legal age. In a system with inadequate documentation of births, such a claim could easily see many adults through. Thus, the punishment stipulated under the law could hardly be a credible deterrent for a determined parent. Consequently, the law is widely thought to have failed in its objective. The 1931 Census report indicated that the Act was a “dead letter” in most provinces (Whitehead 1995). At least one contemporary women’s activist shared the same opinion (Sukathankar 1935). In August 1932, two and a half years after the Act was passed, there were only 173 prosecutions, out of which 167 (more than 95%) were successful. This is evidently the pattern even in more recent times. The number of cases brought to court for all India as a whole was 85 in 2001, 113 in 2002 and 63 in 2003. These are minuscule numbers given that the practice of child marriage is very widespread in large parts of India. On the basis of figures such as these, modern evaluations of the Age at Marriage

Economic & Political Weekly

may 23, 2009 vol xliv no 21

i mpressionistic evidence is mixed.

Could the Sarda Act have been effective in raising the age at marriage simply because law-abiding individuals might have liked to comply with the law? In particular, can the sharp increase in the age at marriage that was observed in our sample for 1930 be attributed to the Sarda Act? We must examine our data to see if the percentage of marriages of girls under 14 years declined substantially in 1930 relative to earlier years. In 1929, only about 12% of the marriages of women in our sample occurred at less than 14 years of age, whereas in 1930, the year in which the Act became effective, the proportion went up to 20%. A strong possibility is that parents were eager to beat the law and hence got their underage daughters married before the Act came into effect. In fact, some social groups carried out large-scale marriages of underage children as a protest which by and large went unpunished (ibid). This implies that the enactment of the Act could not have raised the age at marriage in 1930, rather it was more likely to have lowered it.


This paper examines one aspect of the historical experience of the middle classes in western India, namely, the increase in the age at marriage for women. It is important to examine this variable since it is one of the dimensions distinguishing the middle class from other social groups. It is also an aspect of the larger change emphasising higher human capital accumulation and lower fertility that occurred among the middle classes before other social groups. Since such transitions are important markers of economic d evelopment, it is important to study the reasons and the timing of their occurrence among different sections of society. Our results show that the ideology of nation building might have mattered to



The Department of Economics, Delhi School of Economics invites applicants for its PhD/MPhil/MA courses in Economics for the academic year 2009-10.

PhD/MPhil admissions are based on an entrance test followed by an interview of shortlisted candidates. Some fellowships are also available for the candidates who are admitted. PhD applicants not requiring financial aid may submit a research proposal instead of taking the entrance test.

MA admissions are based on an entrance test only.

The entrance test for all courses will be conducted on Saturday June 27, 2009 at 10 a.m. at the Delhi School of Economics.

The application form can be downloaded from our website: by clicking on the ‘programs’ link. The website also contains information on eligibility criteria and other details about admission.

The last date for receipt of application forms for all courses is 19 June 2009.

Head of the Department


Global Economic & Financial Crisis Essays from Economic and Political Weekly

In this volume economists and policymakers from across the world address a number of aspects of the global economic crisis. One set of articles discusses the structural causes of the financial crisis. A second focuses on banking and offers solutions for the future. A third examines the role of the US dollar in the unfolding of the crisis. A fourth area of study is the impact on global income distribution. A fifth set of essays takes a long-term view of policy choices confronting the governments of the world.

A separate section assesses the downturn in India, the state of the domestic financial sector, the impact on the informal economy and the reforms necessary to prevent another crisis.

This is a collection of essays on a number of aspects of the global economic and financial crisis that were first published in the Economic & Political Weekly in early 2009.

Pp viii + 368 2009 Rs 350

Available from

Orient Blackswan Pvt Ltd

Mumbai Chennai New Delhi Kolkata Bangalore Bhubaneshwar Ernakulam Guwahati Jaipur Lucknow Patna Chandigarh Hyderabad Contact:

may 23, 2009 vol xliv no 21

this development over and above the other more material factors group. Our story will not be complete unless we can furnish these that were driving the age at marriage for women. We still need to reasons. However, at this stage, it seems quite probable that ideoaccount for the reasons for which the nationalist constructs of the logical construction played an important role in fashioning of the “ideal wives and mothers” found such ready reception among this Chitpavan brahmins into a middle class.


1 “For a discussion of women’s participation in the Civil Disobedience Movement”, see Forbes (1998), pp 129-49.

2 For examples, see Heimsath (1965, 1961) and those cited in Forbes (1998). 3 To cite a few examples, see Tamhankar (1929a), Kanitkar (1928), Tamhankar (1929b). 4 For example, see the advertisements appearing in Tikekar (1929), pp 17-18. 5 For a good survey, see Thapar-Björkert, pp 1-53.

6 Patterson (1968), Patterson (1970) uses kulvritantas to study changing patterns of migration among Chitpavan brahmins.

7 The following genealogies have been used in this study: Khare (1940), Khadilkar (1988), Gadre (1943), Ganapule (1953), Gokhale (1978), Gunye (1944), Pendse (1938), Chiplunkar (1943), Jog (1953), Dongre (1943), Datye (1976), Nitsure (1947), Ponkshe (1949), Deodher (1992), Bapat (1965), Apte (1919), Kunte (1982).

8 About 4% of the population of the Bombay Presidency, see Census of India (1931).

9 For the detailed method of estimating mean age at marriage, as well as the estimated numbers, see Hatekar, Mathur and Rege (2007).

10 Enders (2003), pp 213-14. The null hypothesis of a unit root test could not be rejected in the case of the logarithm of the schools variable. Since the sample size is small, and the unit root tests have low power in small samples, KPSS test where the null hypothesis is trend stationarity was carried out (Kwiatkowaski et al 1992). In this case, the null was rejected at 5%. Since the schools variable appears to have a trend break in 1927, a unit root test with a known structural break (Perron 1989) was carried out. The null of a unit root could not be rejected.

11 In the case of the Weibull estimates, the F statistic had a p value of 0.15, while in the case of the arithmetic mean estimates, the p value was 0.77.


Apte (1919): Apt Apte, Pune.

Bains, J A (1931): Census of India: Operations and Results in the Presidency of Bombay Including Sind,

Vol 1 (Bombay: Government Printing Press).

Bapat, V S (1965): Bapat-Kul-Vritanta, Pune.

Bedekar, Malati (1947): “Jata Jata”, Stree, March.

– (1976): Kalyanche Nishwas (Mumbai: Popular Prakashan). Bole, V S (1921): “Grihashiksha”, Bhandari, April, p 18.

Bhat, P N Mari and S S Halli (1999): “Demography of Bride Price and Dowry: Causes and Consequences of the Indian Marriage Squeeze”, Population S tudies, Vol 53(2), 129-48, July.

Caldwell, J C, P H Reddy and P Caldwell (1983): “The Cause of Marriage Change in South India”, Population Studies, Vol 37, No (3): 343-61.

Cameron, Colin A and P K Trivedi (2005): Microeconometrics: Methods and Applications (Cambridge: Cambridge University Press).

Chandra, Bipin, M Mukherjee, A Mukherjee, K N Pani kkar and S Mahajan (1999): I ndia’s Struggle for Independence (New Delhi: Penguin).

Chiplunkar, G B (1943): Chiplunkar Kul Vritanta, Pune. Chiplunkar, G M (1929): “Stri Hriday ani College Shikshan”, Yashwant, April, pp 9-19. Chatterjee, P (1999): The Nation and Its Fragments (New Delhi: Oxford University Press), pp 116-57. Das, G (2000): India Unbound: The Social and Economic Revolution from Independence to the Global Information Age (New York: Anchor Books).

Datar, R (1920): “Hundyachi Ghatook Chaal”, Udyan, April-June, p 176.

Datye, B S (1976): Date-Kul-Vritanta, Pune.

Deodher (1992): Deodhar Kulvritanta, Mumbai.

Devika, J and Avanti Mukherjee (2007): “Re-forming Women in Malayalee Modernity: A Historical Overview” in S Mukhopadhyay (ed.), The Enigma of Kerala Woman: A Failed Promise of Literacy (New Delhi: Social Science Press).

Doepke M and F Ziliboti (2005): “Social Class and the Spirit of Capitalism”, Journal of the European E conomic Association, 3: 516-24.

– (2007): “Occupational Choice and the Spirit of C apitalism”, NBER Working Paper. Dongre (1943): Dongre Kulvritanta, Pune. Enders Walter (2003): Applied Econometric Time S eries (New Delhi: Wiley Eastern). Fischer, Louis (2007): The Life of Mahatma Gandhi (New Delhi: Harper Collins).

Forbes, Geraldine (1998): Women in Modern India, The New Cambridge History of India, IV2 (Cambridge: Cambridge University Press).

Gadre, G M (1943): Gadre Kul Vritanta, Pune.

Ganapule, P G (1953): Ganpule Kul Vritanta, Baroda.

Glick P C, D M Heer and J C Beresford (1967): “Family Formation and Family Composition: Trends and Prospects” in Sussman (ed.), Sourcebook of M arriage and the Family (New York: Houghton Miffin Co), pp 1-38.

Gogte, G S et al (1972): Gogate Kul Vritanta, Mumbai. Gokhale (1978): Gokhale Kulvritanta, Pune. Government of India (1886): Selections from the Records

of the Government of India, Home Department, P apers Relating to Infant Marriages and E nforced Widowhood in India, Serial Number 3, Calcutta.

  • (1925): Report on Public Instruction in Bombay Presidency, 1921-25 (Bombay: Government Central Press), p 86.
  • (1929a): Progress of Education in India 1922-27, Ninth Quinquennial Review, Vol 1 (Calcutta: G overnment of India).
  • (1929b): Report of the Age of Consent Committee, 1928-29 (Calcutta: GoI).
  • (1932); Quinquennial Report on Public Instruction in Bombay, 1927-32, Part 1, p 159. Goyal, M (2003): “Urge to Splurge”, India Today International, 9: 28-37. Gunye, V T (1944): Gun(y)e Gharanyacha Itihas (Vols 1 and 2), Pune.
  • Guttmacher Institute (1987): “Age of Indian Brides Rising, But Many Still Marry at Ages Younger Than the Minimum Set by Law”, International Family Planning Perspectives, Vol 13(4): 148, December.

    Hajnal, J (1953): “Age at Marriage and Proportions Marrying”, Population Studies, 7(2): 111-36, N ovember.

    Heimsath, Charles (1961): “The Origin and Enactment of the Indian Age of Consent Bill 1891”, Journal of Asian Studies, Vol 21(4): 491-504, August.

    – (1965): Indian Nationalism and Hindu Social Reform (Princeton: Princeton U niversity Press).

    Hirose, H (1999): “Bias Correction for the Maximum Likelihood Estimates in the Two-Parameter Weibull Distribution”, IEEE Transactions on Dielectrics and Electrical Insulation, Vol 6, No 1, February.

    Johnson, G (1970): “Elites in South Asia” in Edmund Leach and S N Mukherjee (ed.), Elites in South Asia (Cambridge: Cambridge University Press), pp 95-118.

    Kanitkar, K (1928): “Lakshmine Dileli Goshta”, K irloskar Khabar, November, pp 122-24. Kanitkar, S R (1953): Jog-Kul-Vritanta, Pune. Karve, D K (1929): Atmavroota, Balikashram, Hingne Budruk.

    Kelkar, I (1919): “Stree Hee Surashtra Nirman Karnari Rashtrajananich Hoy”, Manoranjan, February, pp 151-68.

    Kende, Savitribai (1927): “Streeyansathee Vachanalaye” in Kirloskar Khabar, Streeyance Paan, J anuary, p 10.

    Khadilkar (1988): Khadilkar Kul Vritanta, Mumbai.

    Khadilkar Kulvritanta Committee (1988): Khadilkar Kul Vritanta, Mumbai.

    Khare, M G (1943): Dongre-Kul-Vritanta, Pune.

    Khare, M H (1940): Khare-Kul-Vritanta, privately printed, Pune.

    Krishnakumari (1927): “Striyancha Samajeek Darja” in Grihalakshmi, September Issue, p 14.

    Kulkarni, N V (1929): “Grihini Sachivah”, Masik M anoranjan, July-December, pp 408-17.

    Kunte (1982): Kunte Kulvritanta, Pune.

    Kwiatkowski, D, P C B Phillips, P Schmidt and Y Shin (1992): “Testing the Null Hypothesis of Stationary against the Alternative of a Unit Root”, Journal of Econometrics, 54: 159-78.

    Lawless, John F (2003): Statistical Models and Methods for Lifetime Data (New Jersey: Wiley Interscience), pp 218-23.

    Misra, Maria (2007): Vishnu’s Crowded Temple: India since the Great Rebellion (London: Allen Lane).

    Nitsure, G B (1947): Nitsure-Kul-Vritanta, Pune.

    Pathak, G R (1978): Gokhale Kul Vritanta, Pune.

    Patterson, Maureen L P (1968): “Chitpavan Brahman Family Histories, Sources for a Study of Social Structure and Social Change in Maharashtra” in Milton B Singer and Bernard S Cohn (ed.), Structure and Change in Indian Society (Chicago: A ldine), pp 397-411.

    Patterson, Maureen (1970): “Changing Patterns of Occupation among Chitpavan Brahmins”, The Indian Economic and Social History Review, Vol VII, (3): 375-96, September.

    Pendse, K V (1938): Pendse Kul Vritanta, Pune.

    Perron Pierre (1989): “The Great Crash, the Oil Price Shock, and the Unit Root Hypothesis”, Econometrica, (57): 1361-1401, November.

    Ponkshe, B S (1949): Ponkshe-Kul-Vritanta (Pune: Bharat Itihas Sanshodhak Mandal).

    Rohatgi, V K (2003): Statistical Inference (New York: Dover Publications).

    Rhode, D (2003): “Sleepy City Has High Hopes, Dreaming of High Tech”, New York Times, 20 November.

    Srinivas, M N (1952): Religion and Society among the Coorgs of South India (Oxford: Clarendon Press).

    – (1956): “Sanskritisation and Westernisation” in A Aiyappan and A K Balratnam (ed.), Society in I ndia, Madras, pp 73-115.

    Sukathankar, M (1935): “Mahilonnaticha Marg” in Navi Grihalakshmi, June Issue. Tamhankar, N D (1929a): “Bailwedacha Pratap”, K irloskar Khabar, April, pp 247-51.

    – (1929b): “Karan Samajale”, Kirloskar Masik, May, pp 247-51.

    Thapar-Björkert, S (2006): Women in the Indian National Movement: Unseen Faces and Unheard Voices 1930-42 (New Delhi: Sage).

    Tikekar, C R (1929): Doot-Audyogic – Kulwadhuncha Dyankosh, August, pp 13-17.

    University of Bombay (1876): Essay on the Effects of Early Marriages, awarded the Kursandas Muljee Prize for the best essay on the subject, by an u nknown writer, Mumbai.

    Verma, S (2003): “Marriage Squeeze in India”, Demography India, Vol 32(2): 181-199.

    Whitehead, J (1995): “Modernising the Motherhood Archetype: Public Health Models and the Child Marriage Restraint Act of 1929”, Contributions to Indian Sociology (n s), 29, 1&2, pp 188-209.

    may 23, 2009 vol xliv no 21

    Dear reader,

    To continue reading, become a subscriber.

    Explore our attractive subscription offers.

    Click here


    (-) Hide

    EPW looks forward to your comments. Please note that comments are moderated as per our comments policy. They may take some time to appear. A comment, if suitable, may be selected for publication in the Letters pages of EPW.

    Back to Top