SPECIAL ARTICLE

The Golden Cage

Stability of the Institution of Marriage in India
K Srinivasan, K S James

The form and nature of marriage and family life have

changed over the past few decades in Western societies
and in East Asia, but they have taken different pathways.
Reproduction is becoming delinked from marriage in
the West, while in East Asia remaining single has
become more of a norm. Looking at how the various
factors operating in these societies impinge on
marriages in India, this paper finds that while
development has contributed to a significant rise in age
at marriage, it has not altered the ultimate proportion of
the population getting married by 034. These figures
are in stark contrast to what is observed in the West and
Japan. Deeply rooted in religion and caste, and with
marital breakdown facing punishing social and
economic costs, the institution of marriage is strong in
India and unlikely to show signs of a breakdown in the
near future.

The authors are thankful to Vaithilingam, documentation specialist at

the IIPS; N Kavitha and Annie George, ISEC; and Senthil Selvan,
research assistant.
K Srinivasan (ksrini_02@yahoo.com) is with the International Institute
for Population Sciences, Mumbai and National Fellow, Indian Council
for Social Science Research; K S James (ksjames@gmail.com) is with the
Institute for Social and Economic Change, Bangalore.

38

1 Introduction

he institution of marriage, under which reproduction

has been expected to take place across most cultures
and religious groups over the centuries, is found to be
breaking down in Western societies. There is a growing concern
that this may spread across the globe (Inglehart 1970, 1985;
Lesthaeghe 2002, 2010; Weiss 1975, 1979). According to statistics compiled by the United Nations (UN) on marriage (2012), a
very high proportion of women in the age group 3034 reported
they were unmarried in many West European countries and in
the United States (US). In 2009, it was 48.7% in France, 47.8%
in the United Kingdom (UK), and 26.3% in the US. Many of
these women who are unmarried, legally live with partners in
conjugal relationships called live-in arrangements. These are
cohabitations without formal marriage between the sexes and
even persons of the same sex, and they have become an alternative in many Western countries. In 200110, more than 50%
of births in the UK were by unmarried women. And among
married women, divorce rates are rapidly rising. Even in 1979,
there was approximately one divorce for every two marriages
in the US. All these developments have been considered to be
the direct consequences of a demographic transition and it is
believed that all nations will eventually have to experience a
second demographic transition of this nature (Lesthaeghe 2010).
But East Asian countries that have undergone a rapid demographic transition show a different pattern of marriage and
divorce. They have far greater stability of marriage and far
lower incidence of divorce. At the same time, one of the major
changes in these countries is the proportion of never married
men and women. For instance, in the age group 3034, there
were 19% singles in South Korea and 35% in Japan in 2009.
But only 1.6% of children were born out of wedlock in South
Korea in 2007 and the number was very small in Japan (nearly
40% of babies were born outside marriage in the US in the
same year). Though the percentage of singles in Japan and
South Korea is relatively high, there are practically no births
outside marriage in these countries. Even live-in relationships
in Japan were quite low at 2% among women 20 years and
above, compared to 14% in France and 9% in the UK.
The form and nature of marriage and family life have undoubtedly changed over the past few decades in Western
societies and in East Asia, but they have taken different pathways. Reproduction is becoming delinked from marriage in the
West, while in East Asia remaining single has become more of a
norm. Both result in far below replacement-level fertility in
MARCH 28, 2015

vol l no 13

EPW

Economic & Political Weekly

SPECIAL ARTICLE

these countries. Live-in arrangements are widespread, as are

divorces, in the West. While these are still relatively rare in
East Asia, their pace has been picking up in recent times.
A great deal of social science research has focused on
marriage and divorce, and the research shows that some of
the major factors contributing to the breakdown of marriage
in the West have been identified. The second demographic
transition theory has been an outcome of the empirical examination of such trends across countries (Lesthaeghe 2010;
Lesthaeghe et al 2006). To quote Lesthaeghe, During the first
demographic transition (FDT), the decline in fertility was unleashed by an enormous sentimental and financial investment
in the child (that is, the king child era to use Ariss term),
whereas the motivation during the second demographic
transition (SDT) is adult self-realisation within the role or
life style as a parent or more complete and fulfilled adult.
The breakdown of the institution in the West seems to
have commenced in the 1960s and is continuing, while some
postulate that this will eventually happen in Asian countries
as well (Lesthaeghe 2010).
In the available literature, scholars emphasise four major
causes for the breakdown of marriages in the Westthe rise
of individualism; the economic independence of women and
rising cost of marriage for them; skill specialisation; and
movements towards gender equalityall of which were
facilitated by the discovery and easy availability of female
contraceptives.

spheres of employment, education, healthcare, and social

services. In schools and colleges, the courses offered for study
are becoming varied and specialised. It is the same in manufacturing, distribution of goods and services, healthcare,
banking, and other services. Many responsibilities have been
transferred from the family to specialised services such as
homes for the aged, security services, maintenance services,
and the like (Desai and Dubey 2011). Many traditional functions of the family have been taken over by non-familial
institutions as a part of development, and marriage has
become a casualty.

Rise of Individualism

2 Objectives of the Study

The spread of individualism in most aspects of life was observed by many as early as in the middle of the 20th century.
While studying US families, Kuhn (1955) observed that the rise
of individualism was not limited to one cause, but was the
result of several factors, such as Protestantism, capitalism, and
the breakdown of inherited and ascribed relationships.

It is possible that each of the contributory factors that has

operated in the West is not relevant to the Indian cultural
context and the institution of marriage is relatively stable
here. However, fears have been expressed in different quarters
that what began in the West and has extended to other parts of
the world will occur in India as well, and that there are already
signs and symptoms of it (Lesthaeghe 2010; Srinivasan 2014).
A web-based survey of professionals in India was carried out
from October 2013 to January 2014 by the senior author,
mainly to ascertain their views on various population-related
issues, policies, and programmes in the country since 1951.
A question was also asked on what they thought of the
stability of the institution of marriage in India. The survey
covered a sample of 242 professionals, who were just 15% of
those to whom the questionnaire was sent. A sample of
non-respondents was later compiled to get their background
details to assess the differences between distributions of
respondents and non-respondents. Since there were no
significant differences on a number of variables, it was decided to proceed with the analysis.
The analysis showed that more than half the respondents
(53%) perceived that marriage as an institution within
which child bearing should take place would break down in
India and child bearing would eventually be delinked from
marriage. The rest (47%) did not think that this would
happen at any time in the near future. Though this finding
is based on a small sample, which may not reflect the real

Costs of Marriage

Another reason identified is the cost of forming a family.

According to Becker (1960, 1996) and many other economists,
all human decisions, including personal ones, are ultimately
based on an economic cost-benefit analysis, whether done
overtly or not. When a society is economically backward, it
makes economic sense to marry and have children to add to
the labour force of the family, and living in one house reduces
costs. When a society develops and the costs of rearing
children rise, parents control their family size with decisions
of the baby or Baby Austin type. When a society becomes
economically more advanced and women with specialised
skills have good employment potential, the costs of marriage
far outweigh its economic benefits for many of them.
Specialisation of Skills

There is growing specialisation of skills occurring as a natural

by-product of the rapid economic development and growth of
consumerism in the West, and this is happening in developing
countries as well, with a time lag. Specialisation is seen in the
Economic & Political Weekly

EPW

MARCH 28, 2015

vol l no 13

Gender Equality

Most societies in the world have been based on patriarchy for

centuries, with women playing a secondary or subordinate
role to men. This dramatically changed in many parts of the
world in the past century with women becoming more educated,
economically gainfully employed, and politically empowered,
while womens movements argued for gender equality in all
aspects of life. Womens equity and equality have been
demanded as a fundamental part of human rights. Powerful
feminist movements in Europe and the US have written about
and argued for giving women the freedom to shape their own
lives. The institution of marriage has been considered by many
as a hindrance to securing gender equity and equality, and
those aspiring for female autonomy and gender equity view it
as a remnant of the patriarchal system.

39

SPECIAL ARTICLE

picture, it has rung some alarm bells on the future of marriages in India.
This paper looks into the institution of marriage in India
with two objectives in mind. First, it analyses changes in marital
distributions in the country, especially in the proportions of
single, currently married, and previously married women, to
study secular trends in them and explore whether there are
signs of an imminent breakdown, as is happening in the West.
Second, it compares trends in marital status distributions in
India and Japan with those in France, the UK and the US, and
interprets the data on a later age at marriage and a larger proportion of women remaining single till the end of their reproductive period. As marriage in India has deep cultural roots,
and origins and foundations in the caste system, the factors
that contribute to its stability, or lack of it, may be different,
and this is also addressed.
3 Data Used and Methods of Analysis

In this study, which is to be considered as a precursor for more

focused investigations, we analyse the data on marital status
distributions at two levelsthe macro and micro. At the
macro level, we use data compiled from the population
censuses for India and four selected states from 1961 to 2001,
and from the Sample Registration System (SRS) for 2011 (they
were not available from the 2011 Census at the time of this
study). Since the sample used for the SRS is quite large, we feel
that the findings may not be very different from what we
would have if we had used the 2011 Census data on age-sexmarital status distributions.
For the state-level analysis, we selected Bihar, Kerala,
Tamil Nadu, and Uttar Pradesh (UP). Kerala and Tamil Nadu
are considered to be more developed states, socially and economically, while Bihar and UP represent the less-developed
states. In terms of the Human Development Index (HDI)
values estimated for 23 states by the Planning Commission in
March 2014, Kerala was ranked 1, Tamil Nadu 8, UP 18, and
Bihar 21. For the micro-level analysis, we used data collected
at the individual level by the National Family Health Survey-3
(NFHS-3) (200506).
We compiled data on the reported marital status of women
in four developed countriesFrance, Japan, the UK and the
USfrom 1970 to 2010 from data published by the UN, as well
as the singulate mean age at marriage (SMAM) of these countries from 1961 onwards for comparison with that of India.

2024, and even in 2529, implying a rise in the age at marriage.

From the age group 3034 onwards, there is a secular decline.
This is true about states with an advanced demographic
transition and those with an early demographic transition,
implying a rapid rise in the incidence of marriages across the
country. For example, in Kerala, singles in the age group 3034 in
2011 were 6.3%, while in Bihar they were 1.5%. In the same
year, women remaining single in the age group 5054 in
Kerala were 3.4%, compared to 0.00% in Bihar (approximated
to the second decimal).
Table 1: Percentage of Females Remaining Single in Different Age Groups,
India, Bihar, Kerala, Tamil Nadu and Uttar Pradesh, 19612011 (Percentage)
Year

India
1961
1971
1981
1991
2001
2011
Bihar
1961
1971
1981
1991
2001
2011
Kerala
1961
1971
1981
1991
2001
2011
Tamil Nadu
1971
1981
1991
2001
2011
Uttar Pradesh
1961
1971
1981
1991
2001
2011

Age Group
3034
3539

1519

2024

2529

4044

4549

5054

29.2
43.7
55.9
64.3
75.2
87.8

6.0
9.5
14.0
17.0
23.0
37.3

1.9
2.3
3.3
4.2
5.7
12.2

1.0
1.0
1.2
1.8
2.2
4.0

0.7
0.8
0.6
0.9
1.3
1.4

0.6
0.6
0.5
1.0
1.2
1.7

0.5
0.5
0.4
0.7
0.9
0.0

0.5
0.5
0.4
0.8
0.9
0.0

15.6
23.1
35.3
44.2
60.4
86.0

3.0
3.6
5.2
7.1
9.5
27.8

1.3
1.0
1.3
1.6
1.5
5.1

0.9
0.5
0.5
0.9
0.7
1.5

0.7
0.4
0.3
0.4
0.6
0.0

0.6
0.3
0.3
0.7
0.3
0.0

0.6
0.3
0.2
0.5
0.2
0.0

0.6
0.3
0.2
0.8
0.3
0.0

69.6
81.0
85.4
88.5
86.7
93.1

22.7
32.7
40.2
43.4
41.6
48.1

8.0
9.3
12.5
13.9
13.0
14.6

4.5
5.3
5.8
6.2
5.8
6.3

3.2
3.7
3.5
3.9
3.9
3.9

2.9
3.5
3.4
3.7
3.5
2.6

2.2
3.1
2.9
3.0
3.2
2.8

1.8
2.9
3.1
3.2
3.4
3.5

72.7
76.8
81.9
84.3
94.6

17.0
22.9
28.5
34.8
47.9

2.7
4.8
6.8
8.4
15.2

1.2
1.7
2.5
3.0
4.7

0.7
0.8
1.1
1.6
1.3

0.7
0.7
1.1
1.5
1.4

0.6
0.5
0.7
1.1
1.6

0.5
0.4
0.7
1.1
0.0

17.0
26.6
39.0
53.0
72.6
89.9

2.5
3.8
5.8
8.1
16.1
42.6

1.0
1.1
1.0
1.5
3.0
10.8

0.6
0.6
0.3
0.7
1.0
1.6

0.4
0.7
0.2
0.3
0.6
0.0

0.4
0.4
0.2
0.7
0.6
0.0

0.3
0.3
0.1
0.6
0.4
0.0

0.4
0.3
0.2
0.6
0.5
0.0

Source: Computed from census data for various years.

4 Indicators of Changes in Marriage Pattern

The study has used many conventional indicatorspercentage

remaining single, mean age at marriage, and divorce ratefor
understanding changes in marriage patterns in India in comparison to developed countries.
4.1 Single or Unmarried Women in India

Table 1 presents the percentage of females remaining single in

five-year age groups from 15 to 54 in all India, and in the four
selected states. It can be seen that there has been a steady,
secular increase in the values in the age groups 1519 and
40

Figures 1A and 1B (p 41) give a pictorial representation of the

trends of singles in India and the four selected states for the
age groups 2529 and 3034, respectively. We can see from
Figure 1A that the percentage of singles in all the states seem to
converge over time to the Kerala value of 16.35%, and from
Figure 1B that there is a clear rising trend in all the states to the
Kerala level, but the gaps are still quite wide. While Kerala had
5.3% singles in 2011, all the other states had between 1% and
4%, but rising to the Kerala level. Women in Kerala marry
quite late, in their mid-20s, but by 30 to 35 (around 33), most of
them are married. Marriage is as universal in Kerala as in the
MARCH 28, 2015

vol l no 13

EPW

Economic & Political Weekly

SPECIAL ARTICLE
Figure 1A: Single Females in Age Group 2529 in India and Selected States,
19612011
(Percentage)

Percentage Single

16

12

Uttar Pradesh
Kerala

India

8
Tamil Nadu
4

Bihar
0
1961
1971
1981
1991
Source: Computed from census data for various years.

2001

2011

Figure 1B: Single Females in Age Group 3034 in India and Selected States,
19612011
(Percentage)

For Bihar, it rose from 15.9% in 1961 to 17.1% in 1981, and to

21.0% in 2011; for Kerala, from 20.1% in 1961 to 21.8% in 1981,
and to 22.7% in 2011. For Tamil Nadu, the figures were 18.5%
to 20.3%, and to 23.1%, and for UP, 16.0% to 17.3%, and to
22.2%. Thus, the increase in the SMAM values over the 50-year
period 19612011 was 2.6 years in Kerala; 4.6 years in Tamil
Nadu; 5.1 years in Bihar; and 6.2 years in UP. The data show a
rapid convergence of the SMAM values of the other states to
the Kerala level. Thus the less-developed states are catching
up with the advanced ones in their SMAM values. While the
age at marriage in less-developed states has been rising
rapidly, there is a slowdown in the rate of increase in the more
developed states.

4.3 Comparison with Four Developed Countries

Uttar Pradesh
India

4
Tamil Nadu
2

Bihar

1961
1971
1981
1991
Source: Computed from census data for various years.

2001

2011

other states in spite of its higher literacy rate, better health

conditions, and high status of women. It is worth pointing out
that the percentage of women reported single in the 2011 SRS
data are higher in all age groups compared to the 2001 Census
data, and this requires further investigation. However, the
findings on the proportion of singles till the ages of 2529 and
a sharp decline thereafter remain unaffected.
4.2 Singulate Mean Age at Marriage Values

Figure 2: Singulate Mean Age at Marriage in India and Developed Countries

The SMAM computed on the basis of the data on the percentage

of singles, using the standard procedures developed by Hajnal
(1953), are given in Table 2. For India as a whole it increased
from 16.8% in 1961 to 18.7% in 1981 and to 22.22% in 2011.
Table 2: Singulate Mean Age at Marriage in 1549 Age Group in India and
Selected States, 19612011
Selected States

Bihar
Kerala
Tamil Nadu
Uttar Pradesh
India

The data on the percentage of single women in different age

groups in France, Japan, the UK, and US from 1970 to 2010 are
presented in Table 3. The data presents a totally different
picture from that of India. In 2011, singles in the 3034 age
group in India were 4.0%, but they were 48.7% in France,
47.8% in the UK, 26.3% in the US, and 34.5% in Japan in 2009.
It is obvious that a high proportion of the single women in
these four countries in the age group 3034 may be in live-in
arrangements. Even in the age group 5054, singles in these
four countries were 14.4%, 10.2%, 10.0%, and 8.7%, much
higher than Kerala at 3.5% and India at 0.00%.
The scatter plot of the SMAM values of India compared
with the four developed countries is seen in Figure 2. The
SMAM values in these four developed countries are steadily
increasing, implying a steady rise in live-in arrangements. In
Japan, where only a small proportion of births occur to single

1961

1971

1981

1991

2001

2011

15.93
20.05
18.45
15.99
16.84

16.37
21.07
19.61
16.57
17.76

17.08
21.84
20.27
17.28
18.66

17.56
22.27
20.91
18.06
19.26

18.59
21.96
21.41
19.57
20.20

21.02
22.71
23.12
22.25
22.22

Source: Computed from census data for various years.

40

30
SMAM

Percentage Single

Kerala
6

20

10

0
1950

1960

1970

1980

1990

2000

2010

India
Source: United Nations (2012).

US

UK

France

x|

Table 3: Single Females in Different Age Groups in France, Japan, UK, and US, 19702010
Age

1519
2024
2529
3034
3539
4044
4549
5054

France
1970

1970s
Japan
UK
1970
1971

US
1970

93.7
46.1
16.5
10.5
8.9
8.4
8.3
8.1

97.9
71.7
18.1
7.2
5.8
5.3
4.0
2.7

91.3
40.3
13.9
7.8
7.2
7.7
8.3
9.0

88.1
36.3
12.2
7.4
5.9
5.4
5.4
5.7

France
1985

1980s
Japan
UK
1985
1981

US
1980

France
1999

1990s
Japan
UK
1995
1991

97.9
65.0
27.2
14.4
9.4
7.5
6.7
7.0

99.1
81.6
30.6
10.4
6.6
4.9
4.3
4.4

90.8
51.3
22.0
10.8
8.1
3.9
4.8
4.8

99.4
93.1
66.2
40.0
26.2
16.7
11.6
8.5

99.3
86.8
48.2
19.7
10.1
6.8
5.6
4.6

EPW

MARCH 28, 2015

95.5
53.7
19.2
8.8
6.2
5.6
6.0
6.8

2020

Census/ surveyed years

98.2
75.4
38.4
18.2
10.2
6.4
5.2
5.0

Japan

(Percentage)
US
1990

France
2005

2000
Japan
UK
2005
2001

94.3
64.6
32.0
18.2
10.0

99.1
90.8
66.8
44.5
32.4
23.4
16.2
11.5

99.2
88.7
59.1
32.0
18.8
12.2
8.3
6.2

5.6

95.2
69.1
38.1
21.9
14.3
9.8
7.1
5.2

US
2000

France
2009

2009-10
Japan
UK
2010
2009

94.1
69.1
38.1
21.9
13.4

99.5
92.9
70.8
48.7
36.5
27.9
20.7
14.4

99.4
89.6
60.3
34.5
23.1
17.4
12.6
8.7

8.0

99.6
93.3
71.9
47.8
31.8
22.0
15.1
10.2

US
2009

97.2
77.4
46.3
26.3
16.4
13.1
11.5
10.0

Source: United Nations (2012).

Economic & Political Weekly

vol l no 13

41

SPECIAL ARTICLE

mothers, live-in arrangements are not as widespread as in the

Western countries.
4.4 Incidence of Divorces

Figure 3 has a bar diagram of the percentage of women in the

age group 3034 who reported they were divorced in the four
developed countries in 200910 and in India in 2001. The
extent of divorces in India in recent years is not reliably available.
There is a likelihood of more separations than formal
divorces here, and the reported divorce rate could possibly be
an underestimate. But this is probably true of developed
countries as well in recent times. Although most formal
marriages are registered in developed countries, separations
from live-in arrangements are not usually reported as divorces,
irrespective of the duration. Since the duration of live-in
arrangements is increasing in developed countries, actual
rates of divorce may be higher than what is reported. Thus
there is a problem in the quality of data on the percentage of
women reported as divorced, both in India and the developed
countries. But, the order of differences may indicate a picture
not far from reality.
Figure 3: Divorced in Age Group 3034 in India and Developed Countries
8.00

Percentage Divoced

8
5.80

6
4.70

4.50

2
0.38
0

India 2001

France 2009

Japan 2010
Country/Year

UK 2009

US 2009

Source: United Nations (2012).

From Table 3 and Figure 3, it can be seen that even while the
marriage rate is declining, the percentage of women divorced
within this declining group has been increasing over the years.
In France, it was 4.9% in 1970, which increased to 16.3% by
2010; in Japan, it increased from 3.8% to 9.2%; in the UK from
2.0% to 19.0%, and in the US from 5.5% to 17.4%. The rise in
divorce rates is the lowest in Japan.
Table 4: Singulate Mean Age at Marriage in India and Developed Countries
India
US
UK
France
Japan

1961

Circa 1970

Circa 1980

Circa 1990

Circa 2000

16.84

17.76
20.5
21.3
22.3
24.7

18.66
23.3
23
24.7
25.8

19.26
25.4
26.4
30.7
27.7

20.20
26
26.3
31
29.4

Latest

21.75
26.9
31.8
31.6
29.7

Source: United Nations (2012).

4.5 Single Status among Men

When we study the universality of marriage and its stability

in India, it is important to corroborate the findings-based
data on females with similar data on men. Table 5 presents
data on the percentage of single men in 2011 (based on SRS
age-sex-marital status distributions) in India as a whole and
the four selected states. We see that in India single men were
42

97% in the age group 1519, which declined sharply to 14% in

the age group 3034, and to 3% in the age group 5054. In
Kerala, single men in these three age groups were 100%, 27%,
and 2%. The trend of decline in the percentage of singles in
the four states clearly shows a convergence to almost 0%
by age 55.
Table 5: Single Men in India and Selected States, 2001
Age

1519
2024
2529
3034
3539
4044
4549
5054
5559

India

Bihar

Kerala

97.20
74.47
37.78
13.51
4.35
3.39
1.85
2.50
0.00

95.90
65.48
27.85
6.67
1.69
0.00
0.00
0.00
0.00

100.00
94.94
66.67
27.03
9.59
4.17
2.94
1.69
1.82

(Percentage)

Tamil Nadu

Uttar Pradesh

100.00
90.36
54.26
19.32
6.41
2.86
1.56
1.89
0.00

97.64
73.00
34.44
11.11
4.92
4.26
2.27
3.23
3.57

Source: Computed from census data for various years.

5 Findings from Micro-level Analysis

The analysis in the above sections is based on aggregate

data at the state and national levels in India and the selected
developed countries for comparative purposes. They provide
only broad patterns on the distributions of women by marital
status over time. In the micro-level analysis, we used data
collected in the NFHS-3, coordinated by the International
Institute for Population Sciences (IIPS) under the aegis of the
Ministry of Health and Family Welfare, in 200506. In this
survey, all women in the reproductive age range 1549 were
interviewed, unlike in NFHS-1 and NFHS-2 where only ever
married women were considered. The NFHS-3 surveyed a
nationally representative sample of 1,09,041 households, 1,24,385
women aged 1549, and 74,369 men aged 1554, and this
covered 99% of the countrys population in all the 29 states.
We analysed the data on the marital status of women
reported in the survey, grouping them as never married/single;
ever married (includes currently married); and previously
married (includes widowed, divorced, and separated). We
used a simple binomial logit regression using five predictors
age, rural/urban status, education level, wealth index of the
household, and the variables of religion and caste combined as
a religio-caste variable. While age is used as a continuous
variable, the other four are used as discrete or categorical
variables. For religio-caste, we used a single ascribed variable
that combines religion and caste among Hindus. We used
dummies for the four states. The categorisation of the religiocaste variable that we used was(i) Christians; (ii) Muslims;
(iii) Hindu Scheduled Caste (SC); (iv) Hindu Scheduled Tribe
(ST); (v) Hindu Other Backward Class (OBC); (vi) Hindu others;
and (vii) Other religions. The reason for combining religion
and caste into a single variable was that caste, which was predominantly a nomenclature assigned at the time of birth to a
child born in the Hindu religion, is now increasingly prevalent
in other religious categories. We found that caste categorisation among religions other than Hindus confounds the effects
of caste and religion on marital status, and this requires
introducing a new variable.
MARCH 28, 2015

vol l no 13

EPW

Economic & Political Weekly

SPECIAL ARTICLE

From Table 6, it can be seen that all the variables selected

as predictors have highly significant odds ratios on the probability of getting married. Age, used as continuous variable,
has an odds ratio of 1.5, and indicates that, controlling for
other factors, it increases the odds of marrying by 50%. Even
controlling for age and other achieved variables such as place
of residence, education, and wealth, the religio-caste variable
has the most effect on Hindus in all caste categories, with an
odds ratio more than 2, compared to Christians. The highest
odds ratio of 2.94 is found among Hindu OBCs. The odds ratios in
all categories of the three achieved variables of rural/urban
residence, education of women, and wealth index are under 1,
indicating that the maximum probability of marriage is among
rural, illiterate, and poor women, and that changes in these
variables reduce the odds ratios, contributing to upward mobility. The most important finding is the high odds ratios of the
religio-caste variable. Using state dummies in the logit reveals
that Tamil Nadu (0.55) and Kerala (0.85) have odds ratios below 1, while Bihar (1.57) and UP (1.04) have above 1, with all
the other states combined used as the base for comparison.
Table 6: Logistic Regression Result on Odds of Being Married, 200506
(0 = unmarried; 1 = ever married)
Background Characteristics
Coefficient

Type of place of residence

Age of women
Religion and caste

Educational level of women

Wealth index

State

Rural
Urban
Current age
Christians
Muslims
Other religions
HinduSC
HinduST
HinduOBC
HinduOthers
Illiterates
Primary
Secondary
Higher
Poorest
Poorer
Middle
Richer
Richest
All other states
Bihar
Kerala
Tamil Nadu
Uttar Pradesh

Constant

India
Odds Ratio

.401
.407

.670***
1.503***

.601
.295
.983
.700
1.078
.733

1.823***
1.343***
2.674***
2.014***
2.937***
2.082***

.587
1.409
2.819

.556***
.244***
.060***

.014
.177
.234
.398

.986***
.837***
.791***
.672***

.454
.168
.594
.041
7.263

1.574***
.846***
.552***
1.042***
.001

Source: Estimated from National Family Health Survey-3 (200506).

The main inference from this analysis is that the ascribed

variables of religion and caste are important factors affecting
the probability of marriage for single women even after controlling for the achieved variables of rural/urban residence;
education of women, and index of family wealth. As we will be
arguing, the institution of marriage has, over the centuries,
been strongly embedded in a context of religion and caste.
Inter-caste marriages and inter-religious marriages are socially
Economic & Political Weekly

EPW

MARCH 28, 2015

vol l no 13

condemned and frowned upon. So we find out whether these

two ascribed factors dominate other achievement factors in
determining marriage and age at marriage.
6 Summary and Discussion

In Section I, we listed four major factors that have contributed

to the breakdown of the institution of marriage in the West.
We argue that none of these factors is relevant to India as
things stand now. The rise of individualism in the Western
sense is unlikely to happen in India in the near future since all
individuals in Indian culture are a part of a larger network of
family, gotra, caste, and religion. A girl child is brought up as a
daughter to be married off at a later age, fully dependent on
her parents till marriage, later dependent on her husband, and
if unfortunately widowed, dependent on her son. A boy is told
from a very young age of his responsibilities to his family, parents, and sisters, and the various ills that will descend on him
if he ignores them in their old age. There appears to have been
no major changes in such norms in India even with considerable demographic and socio-economic changes. No religious
groups are an exception to this. One example of this is the
dowry system, which is practised across all religious groups in
India. So there is little chance of anyone born in India declaring that he or she is an individual bereft of all familial connections and concentrating solely on his or her own personal
development. This is not likely to happen, at least for many
decades to come.

Subscribe
to the
Web Edition/
Digital Archives

What do you get with a Web subscription?

t
50 online issues every year
t
Web Exclusives
t
All special and review issues
t
A host of other features on
www.epw.in

t
Access to full archives

from 1949 (including the
Economic Weekly)
t
Featured themes articles
on contemporary issues from
the archives

To subscribe, visit: www.epw.in/subscribe.html

Attractive rates available for individuals and institutions.
Postal address: Economic and Political Weekly,
320-321, A to Z Industrial Estate, GK Marg, Lower Parel, Mumbai 400 013, India.
Tel: +91-22-40638282 | Email: circulation@epw.in

43

SPECIAL ARTICLE

The second factor is the costbenefit analysis of girls marrying within and outside the same caste. In India, many studies
have brought out the economic and emotional benefits of a girl
marrying within the same caste. In an interesting study based
on advertisements that appeared in the matrimonial columns of
Anandabazar Patrika (a Kolkata daily), Banerjee et al (2013)
analysed data from a large number of marriages that had taken
place over many decades through these ads. It included a detailed
econometric analysis of the data collected from the women interviewed, and concluded that marrying within the same caste
is socially and economically advantageous both to the bride and
the bridegroom. To quote, One of our key empirical findings is
that there is a very strong preference for within-caste marriage.
However, because both sides of the market share this preference
and because the groups are fairly homogeneous in terms of the
distributions of other attributes, in equilibrium, the cost of
wanting to marry within-caste is low. This allows caste to remain
a persistent feature of the Indian marriage market.
The extent of same-caste and inter-caste marriages in India
was studied by Das et al (2010), using data from the NFHS-3.
The caste categories used in this analysis were only fourSC,
ST, OBC, and others. Only when a husband and wife belonged
to a category other than these four was it considered an intercaste marriage. Such a broad caste categorisation can grossly
underestimate the magnitude of inter-caste marriages that
take place within each broad category. However the order of
magnitude of inter-caste marriages found in this study is quite
revealing. As a whole, married women in the age range 1549
reporting marriage within the same caste was 89%. In Bihar, it
was 89%; in UP 88%, in Kerala 80%, and in Tamil Nadu 97%
(Das et al 2010). It is really surprising that the figure is so high
in Tamil Nadu, which has had a strong Dravidian movement

for more than six decades and where Dravidian parties that have
officially promoted and rewarded inter-caste marriages have
been in power for more than four decades. Though the quality
of data on caste reported in the NFHS-3 can be questioned, the
magnitude of inter-caste marriages reported across all Indias
states, more developed or less developed, is very low.
The third factor that has contributed to the decline of marriages in the West is over-specialisation of skills. In India, with
a large percentage of workers still employed in agriculture,
this is not likely to happen in the near future. The level of
specialisation of skills observed in the West is not found even
in the secondary and tertiary sectors. And in the IT sector,
most of the jobs are really those outsourced from developed
countries because of differences in the cost of labour.
India is also far from the fourth factor of gender equality
and parity. Based on various measures on gender disparity
such as female feticide, education, employment, freedom to
marry, domestic violence, employment, economic freedom,
and many others, Indian women rank very low compared to
women not only in developed countries, but also in Asian ones.
The poor status of women is reflected in the highly skewed sex
ratio (06) among children, high maternal mortality rates,
high level of malnutrition, and high morbidity rates.
According to the Global Gender Gap Report released by the
World Economic Forum (WEF) in 2011, India was 113 on the
Gender Gap Index (GGI) among 135 countries polled. It
improved its ranking to 105 among 136 countries in 2013, but
this is still quite low in the comity of nations. Various other
gender empowerment measures have scaled India close to
130140 out of 175. In spite of various efforts made by governments, non-governmental organisations, and a few political
parties, gender equalities on a scale achieved in the West will

The Adivasi Question

Edited By

INDRA MUNSHI
Depletion and destruction of forests have eroded the already fragile survival base of adivasis across the country, displacing an
alarmingly large number of adivasis to make way for development projects. Many have been forced to migrate to other rural
areas or cities in search of work, leading to systematic alienation.
This volume situates the issues concerning the adivasis in a historical context while discussing the challenges they face today.
The introduction examines how the loss of land and livelihood began under the British administration, making the adivasis
dependent on the landlord-moneylender-trader nexus for their survival.
The articles, drawn from writings of almost four decades in EPW, discuss questions of community rights and ownership,
management of forests, the states rehabilitation policies, and the Forest Rights Act and its implications. It presents diverse
perspectives in the form of case studies specific to different regions and provides valuable analytical insights.

Authors: Ramachandra Guha Sanjeeva Kumar Ashok K Upadhyaya E Selvarajan Nitya Rao B B Mohanty
Pp xi + 408 Rs 695
ISBN 978-81-250-4716-2
2012

Brian Lobo K Balagopal Sohel Firdos Pankaj Sekhsaria DN Judy Whitehead Sagari R Ramdas Neela Mukherjee
Mathew Areeparampil Asmita Kabra Renu Modi M Gopinath Reddy, K Anil Kumar, P Trinadha Rao,
Oliver Springate-Baginski Indra Munshi Jyothis Sathyapalan Mahesh Rangarajan Madhav Gadgil Dev Nathan,
Govind Kelkar Emmanuel DSilva, B Nagnath Amita Baviskar

Orient Blackswan Pvt Ltd

www.orientblackswan.com
Mumbai Chennai New Delhi Kolkata Bangalore Bhubaneshwar Ernakulam Guwahati Jaipur Lucknow Patna Chandigarh Hyderabad
Contact: info@orientblackswan.com
44

MARCH 28, 2015

vol l no 13

EPW

Economic & Political Weekly

SPECIAL ARTICLE

not be realised in India in the near future. Indian women tend

to be valued by society in relation to their role in the family
as a wife, daughter-in-law, and mother. Women who fall outside these roles, such as widows and single women, face discrimination and, in many cases, loss of property. Since a woman
is considered incomplete without marriage, unmarried adult
women, widows, and divorcees face strong social stigma.
Desai (1994) has a detailed discussion on gender inequalities
and demographic behaviour in India. In this, she argues that
marrying and having a son as early as possible is the only
choice available to most women in India for social acceptability and upward mobility. Thus, in a way, marriage seems to be
the only saving grace for the status of women in Indian society.
The percentage of inter-caste marriages is low equally among
the more educated and among the poorly educated; among
groups economically better off and worse off; and among those
exposed to the mass media and not exposed to it. Marriages
within the same religion were 85% among Hindus, 89% among
Muslims, and 90% among others. It was almost constant at 89%
among those with different standards of living and among those
exposed to the mass media or not. Thus there appears to be a
rigid caste structure that cuts across education and economic
groups influencing marriages in India. It can be inferred that as
long as the caste system is not basically disturbed, the institution of
marriage as it prevails now will not undergo any radical changes.
Unchanging Proportion

Based on the proportions of women in different age groups remaining single and trends over the past 50 years in India as a
whole, and in two more advanced and two less advanced
states, it is found that while development has contributed to a
significant rise in age at marriage, it has not altered the ultimate
proportion getting married by ages 3034. Most women, 97%,
get married by the age of 4549 even in the most developed
state of Kerala, and this is more than 98% in the other three
states. The percentage reporting divorce or separation is very
low at all ages, less than 1%. These figures are in total contrast
to what is observed in the West and Japan. Comparing the levels
References
Banerjee, Abhijit, Esther Duflo, Maitreesh Ghatak
and Jeanne Lafortune (2013): Marry for
What? Caste and Mate Selection in Modern
India, American Economic Journal: Microeconomics, Vol 5, No 2, pp 3372.
Becker, Gary S (1960): An Economic Analysis of
Fertility in National Bureau of Economic
Research (eds), Demographic and Economic
Change in Developed Countries, New York:
Columbia University Press, pp 20940.
(1996): Accounting for Tastes, Cambridge, MA:
Harvard University Press.
Das, Kumudini, Kailash Chandra Das, Tarun
Kumar Roy and Pradeep Kumar Tripathy
(2010): Inter-caste Marriages in India: Has
It Really Changed Over Time?, presented
at the European Population Conference,
Vienna, http://epc2010.princeton.edu/papers/
100157.
Desai, Sonalde (1994): Gender Inequalities and
Demographic Behavior in India, New York: Population Council.
Desai, Sonalde and Amaresh Dubey (2011): Caste
Economic & Political Weekly

EPW

MARCH 28, 2015

and trends in single, married, and divorced women in different

ages in France, the UK, and the US, and Japan with those in India
and four of its states, we conclude that the institution of marriage
is strong in India, and is neither showing nor likely to show (as
feared by some professionals in a survey) signs of a breakdown
in the near future. Micro-level analysis using NFHS-3 data show
age shifts in marital status from single to married overriding
all other factors. The ascribed factors of religion and caste
have more effect on the odds ratios of the probability of getting married than education or wealth status of a household.
A strong contributing factor to the stability of the institution
of marriage is that the bulk of marriages take place within the
same caste, and there is strong caste support to every marriage. This is rather unfortunate because the social evils and
economic impediments of a rigid caste structure are well
known, and many leaders, including M K Gandhi, were against
it. The efforts of many social and political movements, such as
the Dravidian movement in Tamil Nadu, do not appear to have
had any impactonly about 3% inter-caste marriages were reported in the state in the NFHS-3. Analyses of data on requirements for a spouse in advertisements in matrimonial columns
of newspapers and websites show that the same caste is an essential condition for most, irrespective of socio-economic
background. Caste-based vote bank politics has only strengthened the caste system. Perhaps, one of the contributions of the
caste system is the stability it has brought to the institution of
marriage in India. The costs and benefits of the caste system
require a Beckerian analysiswhether disruptions at the family level by increased live-in arrangements, high divorce rates,
and children reared by single parents will be the trade-off for
achieving a more egalitarian and faster going nation. However, women in India, as of now and in the near future, seem
to be safely trapped in a golden cage of marriage propped up
by religion, caste, and economic forces. The stability of the institution of marriage in India is taken for granted by the scholars and the public at large, but this study highlights that the
roots of such stability are in religion, caste, and the high social
and economic costs of marital breakdown.

in 21st Century India: Competing Narratives,

Economic & Political Weekly, Vol 46, No 11.
Eurostat Data Base (2011): Marriage and Divorce
Statistics, European Commission, http://epp.
eurostat.ec.europa.eu/statistics_explained/index.php/Marriage_and_divorce_statistics.
Hajnal, John (1953): Age at Marriage and Proportions Marrying, Population Studies, Vol 7,
No 2, pp 11136.
IIPS and Macro International (2007): National
Family Health Survey (NFHS-3), 200506:
India, Vol 1, Mumbai: International Institute for
Population Sciences.
Inglehart, R (1970): The Silent Revolution, Princeton: Princeton University Press.
(1985): Aggregate Stability and Individuallevel Flux in Mass Belief Systems: The Level of
Analysis Paradox, American Political Science
Review, Vol 79, pp 97116.
Kuhn, M H (1955): American Families Today: Development and Differentiation of Types in
Becker R Hill (ed), Family Marriage and Parenthood, Boston: D C Heath.
Lesthaeghe, R (2002): Meaning and Choice: Value
vol l no 13

Orientations and Life Course Decisions, NIDICBGS Monograph No 37, The Hague: Netherlands Interdisciplinary Demographic Institute.
(2010): The Unfolding Story of the Second Demographic Transition, Population and Development Review, Vol 36, No 2, pp 21152.
Lesthaeghe, Ron J and Lisa Neidert (2006): The
Second Demographic Transition in the United
States: Exception or Textbook Example?,
Population and Development Review, Vol 32,
No 4, pp 66998.
Maslow, A (1954): Motivation and Personality, New
York, Harper and Row.
Srinivasan, K (2014): Indias Population Issues and
Policies: Professionals Perceptions, findings
from a web-based survey submitted to the Indian
Council for Social Science Research (ICSSR).
United Nations (2012): World Marriage Prospects
2012, New York: United Nations, http://
www.un.org/esa/population/publications/
WMD2012/MainFrame.html
Weiss, R S (1975): Marital Separation, New York:
Basic Books.
(1979): Going It Alone, New York: Basic Books.

45