The Effect of Status of Women on Fertility in an Urban Setting of Tamil...
The Effect of Status of Women on Fertility in
an Urban Setting of Tamil Nadu
This paper throws light on the status of women (at the household level) and fertility
linkage with data drawn from 300 currently married women residing in a small
town of the state of Tamil Nadu, India. Cross-tabular, hierarchical and multiple
classification analyses have been used. Results, suggest that all the dimensions of
the status of women have played a crucial role in influencing their cumulative
fertility (children ever born).
Dr. N. Audinarayana is Reader, Department of Population Studies, Bharathiar
University, Coimbatore.
Several empirical studies in the field of population science research
have confirmed that improvement in various dimensions of the status
of women reduces the level of fertility, maternal mortality, infant and
childhood mortality and promotes rapid implementation of family
planning programme ( Mahadevan, 1989; Mason, 1984; 1995.). In
view of this fact, enhancement of women's status has emerged as an
important policy objective in the development plans and population
policies of many less developed countries.
In demography, however, the role of the status of women did not
receive adequate attention till recently. Most of the empirical re-
searches on the status of women-fertility relationship dealt with the
'proxies' (Mason, 1995) namely, women's age, age at marriage,
education, employment/occupation and age difference between hus-
band and wife. Thenmozhi (1993) has proposed a framework for
studying the status of women-fertility relationship wherein these proxy
measures have been taken into consideration as Status Affecting
Variables (SAVs) and are thus highlighted as the factors contributing
to raise women's status at the family level. Of course, it has been
widely discussed and argued by researchers that the status of women

The Effect of Status of Women on Fertility 543
cannot be described or expressed by a single quantitative measure and
is thus multidimensional in nature. It also highlights that the women's
status (autonomy) at the household (familial) level is the crucial one
which will exercise a greater bearing on their reproductive behaviour.
In view of this background, this paper makes an attempt to examine
and discuss the effect of the selected components of the status of
women (at a familial level) on their fertility behaviour.
Theoretical and Empirical Evidence
In the present context, the status of women has been defined in terms
of their autonomy in certain personal, material and domestic affairs.
These include: consultation of women for finalising their marriage;
women's control over jewels brought from natal family; extent of
sex-segregated interaction; extent of restrictions imposed by husbands
on women to do certain tasks; and the extent of women's participation
in decision-making on household matters. Of course, strictly speaking,
all these five are not separate but interdependent components of
women's status.
Consultation of "Women for Finalising their Marriage
In the Indian context, arranged marriages are quite common and,
thereby, the brides are rarely consulted in selecting their life partners
or in finalising their marriage. Thus, the system of arranged marriage
takes away all the freedom of two concerned partners, especially that
of the bride. Such lack of (or little) autonomy of women on the one
hand leads to early age at marriage (Audinarayana, 1990; 1993) and
on the other hand, it encourages the subordinate status of women
(Dubey and Bardhan, 1972) at the familial level even after marriage,
which in turn may increase their level of fertility.
Control over Jewels Brought from Natal Family
Women's control over economic and material resources is considered
to be one of the basic aspects of female autonomy (Dixon, 1978). In
India, where a patriarchal society exists, women do not get a share in
immovable property such as house or land. However, it is of general
practice in this part of the country that women are sent to their in-laws'
place with gold jewellery, mostly as part of dowry or at times as gift(s)
from the natal family. Nowadays, it has become a status symbol too.
But once the woman reaches the husband's place (after marriage),
control over such jewellery may not likely to be in the hands of the

544 N. Audinarayana
concerned woman, because, in most cases, the husband holds (con-
trols) the purse and even the control in other economic/material
resources of the family. In such a situation, the woman experiences
subordinate status, which is likely to lead to higher fertility.
Extent of Sex-segregated Interaction
Interaction of women with men is not that common in the Indian
society. The cultural/traditional codes followed by women namely,
taking meals only after the husband/all other male members of the
family; absence, or negligible participation in the discussions when
husband/male members are discussing either amongst themselves or
with close male relatives/friends; negligible help of husband in house-
hold chores; not accompanying husbands either for shopping or social
functions and so on, reflects the sex-segregated interaction from both
within and outside the family. In a way, these practices indicate the
physical and interactional segregation of the sexes through observation
of purdah or female seclusion (Basu, 1989; Dubey and Bardhan, 1972;
Dyson and Moore, 1983). This shows the women's inequality with
men and the lower status of women, which may exercise a greater
bearing on their fertility.
Extent of Restrictions Imposed on Women by Husbands to do Certain
Activities (Tasks)
With the dominance of a patriarchal and patrilocal society in India,
within the family, women are mostly in a subordinate position, which
may be expressed in a variety of restrictions imposed by husbands on
their wives in performing certain activities (Mukherjee, 1974). These
include: going to shopping and films; spending money on women's
giving clothes, cosmetics, and so on; entertaining their friends/guests;
and giving financial assistance to their relatives. Majority of these
activities are usually not allowed by husbands or are sometimes
allowed only with proper permission, which indicates their personal
autonomy (Visaria, 1993) at the familial level.
Extent of Women's Participation in Decision-making on Household
Majority of Indian women have a greater say in matters like prepara-
tion of food, care of children and home, though they do not have the
necessary power and authority in all spheres of home-making. This
also reflects their subordinate role in the familial level and this, in turn,

The Effect of Status of Women on Fertility 545
is likely to have a greater bearing on their level of fertility. With this
assumption in mind, the respondents have been asked to what extent
they have been involved (participated) in decision-making related to
certain day-to-day household matters and based the responses to these
items an index has been calculated.1
Empirically, in India, a few studies have taken into consideration
one or the other and/or at times all the above components of the status
of women (autonomy), and examined their role on influencing the
fertility (Audinarayana and Thenmozhi, 1992; Dyson and Moore,
1983; Jejeebhoy, 1991; Mukherjee, 1974; Sundar, 1989; Thenmozhi,
1993; Visaria, 1993; Vlassoff, 1992). Most of these studies have
supported the fact that the higher the level of female autonomy (status)
at the familial level, lower will be the fertility and vice versa at different
levels of significance.
For the present study, all these dimensions of the status of women,
have been taken into consideration as predictor variables in determin-
ing fertility besides the duration of marriage as a covariate.
Data Source and Variables Used
Data used for the present analysis have been taken from the study
entitled 'Women's Status and Fertility in an Urban Setting of Tamil
Nadu', carried out by the author during 1991-93. Information was
collected from 300 (randomly selected) currently married couples —
both husbands and wives (wives aged between 15-44 years)— of
Sulur Town of, Coimbatore District, Tamil Nadu, who had at least one
living child at the time of survey.
To examine the effect of the selected dimensions of the status of
women (at the familial level) on fertility behaviour, the following
variables have been used.
Dependent Variable
Children Ever Born (CEB) Actual Number of Live Births
Duration of Marriage Actual Completed Years
Predictor Variables

546 N. Audinarayana
Consultation of Women for Categorised as Not Consulted
Finalising their Marriage and Consulted
Control over Jewels Brought Categorised as Not Brought Jew-
from Natal Family els, No Control and has Control
Extent of Sex-segregated Pooled Score Categorised as
Interaction (Index) Higher (12), Moderate (13-16)
and Lower (17+)
Extent of Restrictions Imposed Pooled Score Categoriesed as
on Women by Higher (9), Moderate (10-13)
Husbands to do Certain and Lower (14+)
Activities (Index)
Extent of Women's Pooled Score Categorised as
Participation in Decision-making Lower (27), Moderate (28-33)
on Household and Higher (34+)
Affairs (Index)
Techniques Used for Analysis
For the purpose of analysis, the differentials in CEB are examined,
through bivariate (cross-tabular) analysis, across the selected dimen-
sions of status of women controlling for duration of marriage. In the
next stage, the multivariate techniques of hierarchical analysis and
Multiple Classification Analysis (MCA)2 have been used to identify
and examine how strongly (in terms of quantitative assessment) each
of these selected dimensions of the status of women has contributed to
the variation in fertility.
Socioeconomic Profile of the Sample Women
The data presented in Table 1 indicates that an overwhelming propor-
tion of sample women (94 per cent) are Hindus and a greater majority
belong to the backward class (71 per cent) as per the Tamil Nadu
Government Order on classification of castes. The sample women
seem to be better educated. While the percentage of illiterates is only
11, eight per cent were educated upto primary level (1-4 standards)
and around 28 per cent upto middle school (5-8 standards) level.
Interestingly, majority of them (44 per cent) have studied upto high
school/higher secondary school level (11-12 standards) and about
one-tenth (nine per cent) have completed college education.

The Effect of Status of Women on Fertility 547
Socioeconomic Profile of the Sample Women
In terms of their work status they are somewhat lagging behind. For
example, a large majority (63 per cent) of them are not participating
in any income generating activities either within or outside the home.
On the other hand, one-fifth (21 per cent) of them are engaged in the
unorganised sector, that is, as labourers (including agricultural), serv-
ice workers (maidservants, and so on) and petty business. The remain-
ing one-sixth (16 per cent) are working in the organised sector of
employment, that is, as textile mill workers, clerical and related work,
teachers, and so on. Thus the sample women are not well placed in
economic pursuits, in spite of living in urban areas, where the avenues
for employment are comparatively higher. Economically, the sample
women are moderately better off. For instance, while 42.7 per cent
belonged to a monthly family income bracket of Rs. 1,500 or less, 30

548 N. Audinarayana
per cent belonged to the Rs. 1,501-2,500 income group and the
remaining 27.3 per cent to higher income group, that is, Rs. 2,501 and
above. The average family income of the sample women is Rs. 1,840.
On the whole, socioeconomically, the sample women may be placed
at a moderate level and thereby, their status would be expected to be
at moderate to higher level in their respective families.
Mean Children Ever Born of Women by Various Dimensions of Status of
Women and Duration of Marriage

The Effect of Status of Women on Fertility 549
Results and Discussion
Children Ever Born: Bivariate Analysis
Table 2 presents the results based on the bivariate analysis about the
average number of children ever born by various dimensions of
women's status controlling for the duration of the marriage. At the
outset, these results highlight that all the five components of the status
of women namely, consultation of women for finalising their marriage,
women's control over jewels, extent of sex-segregated interaction,
extent of restrictions imposed on women by husbands to do certain
tasks and extent of women's participation in decision-making on
household matters have explained significant (at 0.001 level) vari-
ations in CEB and also in the expected direction. Further, it may also
be seen that the differentials in CEB are considerable across the
duration of marriage categories. However, the point to be borne in
mind here is that, through this analysis it is only possible to examine
the gross effects of each of the independent (predictor) variables on
CEB inclusive of the effects that may be attributed to other variables.
Therefore, in the next section, multivariate statistical techniques have
been applied for the same data.
Children Ever Born: Hierarchical Analysis of Covariance
Table 3 provides the results of the hierarchical analysis of covariance
of the children ever born on their duration of marriage as covariate,
and the other five dimensions of status of women, which are listed in
the earlier section, as predictor variables.
The hierarchical analysis of covariance results indicate that all the
predictor variables (including the covariate) together have explained
47.2 per cent variation in CEB. Further, it also highlights that
'women's control over jewels' contributes 10.5 per cent of the total
variation in the CEB (significant at 0.001 level). The 'extent of
women's sex-segregated interaction', and 'consultation of women for
finalising their marriage' have also exhibited a high predictive power
(7.0 and 4.8 per cent, respectively) in explaining significant (at 0.001
level) variation in CEB. On the other hand, though the 'extent of
women's participation in decision-making' and the 'extent of restric-
tions imposed on women by husbands' have shown gross independent
effects on CEB (through bivariate analysis), controlling for other
preceding variables used in the analysis, their role in influencing the
CEB is observed to be at moderate (significant at 0.005 level) and low

550 N. Audinarayana
(significant at 0.010 level) respectively. As expected, the covariate —
marital duration — explains the highest variation (22.9 per cent out of
47.2 per cent) in fertility.
Hierarchical Analysis of Covariance of Children Ever Born on Various
Dimensions of Status of Women
Children Ever Born: Multiple Classification Analysis (MCA)
Table 4 gives the information on the mean number of children ever
born to women under study in the following two forms — unadjusted
and adjusted — for variation accounted for by all the independent
variables and covariate. Along with the category of means, the table
also provides an Eta value, which is a common correlation ratio
associated with the set of unadjusted category effects of each category.

The Effect of Status of Women on Fertility 551
The Eta i indicates the proportion of variance as explained by a
given predictor variable (all categories combined). Associated with the
adjusted category effects for each independent variable, there is a
partial correlation ratio, which is represented by Beta in the table.
These Beta (6) values can be valued as standardised partial regression
coefficients. Finally, the MCA provides a multiple 'R' value, which is
a multiple correlation between the dependent variable and all factors
and the covariate.
Results of Multiple Classification Analysis

552 N. Audinarayana
The MCA results reveal that 'women's control over the jewels
brought from natal family' has a substantial positive influence on the
number of CEB. The unadjusted values show that those women who
have control over the jewels brought from their natal family (through
marriage) had borne conspicuously less number of children (1.68) than
those women who did not have such control (2.36) and had not brought
jewels (2.20). Further, the values of the common correlation ratio
and the partial correlation ratio , associated with this variable in
determining fertility, are observed to be 0.41 and 0.30 respectively.
This indicates that even after making an adjustment for all the variables
and covariate used in the model, the explanatory power of 'women's
control over jewels', in determining the CEB is, more or less, constant
and very high. The table further shows that the variation in CEB is
clear across the categories of their predictor variable. Thus the results
indicate that women's independent control over the jewels brought
from their natal family would enhance their status because of their
economic accessibility and thereby motivate them to curtail the num-
ber of children.
The unadjusted category of means reveal that the 'extent of
women's sex-segregated interaction' has a significant positive in-
fluence on their fertility that is, lesser the sex-segregated interaction
higher the women's status and in turn lesser the number of children.
The average number of CEB to those women who have been
categorised as higher, moderate and lower sex-segregated interac-
tion are found to be 2.50, 1.87 and 1.74, respectively and the value
of 'n' for this variable is as high as 0.43. However, after making
adjustments for the other dimensions of the status of women and
covariate, the prediction power of women's sex-segregated interac-
tion reduced considerably. Similarly, the influence of the 'extent of
restrictions imposed on women by husbands to do certain tasks' and
the 'extent of women's say in decision-making in household mat-
ters', exhibited a higher influence on CEB (unadjusted). After being
adjusted to other variables and covariate, their predictive power has
been reduced substantially. In general, these results highlight the
fact that lesser the restrictions on women by the husband and more
the women's involvement in decision-making in household matters
(which would enhance their status in their respective families and
thereby take a decision to have lower number of children and act
accordingly) lesser the number of children.

The Effect of Status of Women on Fertility 553
As far as the influence of the 'consultation of women for finalising
their marriage' on CEB is concerned, the unadjusted means showed a
clear negative relationship between these two variables that is, those
women who were consulted at the time of their marriage would exert
a higher status which, in turn, lead to lower number of children. The
common correlation ratio (r|) is noted to be 0.32. However, after
making allowances for all the other predictor variables plus the covari-
ate, the relationship between these two variables turned out to be very
weak (6 = 0.03).
Conclusions and Implications
From the foregoing analysis and discussion it is explicit that the various
dimensions of the status of women at the familial level, in general,
have played a crucial role in influencing the cumulative fertility. In the
light of these findings it may be suggested that programmes and
policies have to be formulated and implemented to raise the status of
women at the familial level. This has to be done mainly through the
eradication of social customs, beliefs and traditions which undermine
the value or importance of women in the family and at the societal level
at large. The establishment of voluntary organisations, social educa-
tion to young girls and mass media communication should be directed
against arranged/forced marriage by parents, discrimination shown
towards females (women's subordinate status) both inside and outside
household, and against confining women within the four walls of
Women, in general, may be encouraged to attain higher education
and to participate in economic activities, especially in the modern
sector. Such programmes motivate women to have greater exposure
and interaction outside the familial world, and be economically inde-
pendent. This would enhance their status at home, thereby granting
their greater autonomy in decision-making on their marriage as well
as about their domestic affairs. All these measures, in the long run, will
facilitate fertility decline.
The author wishes to express his sincere thanks to Prof. P.C. Saxena and Dr. C.P.
Prakasam, International Institute of Population Sciences, Mumbai, for their help in
statistical analysis.

554 N. Audinarayana
Extent of Sex-segregated Interaction (Index): 1. Taking meals with husband or other
male family members (Take lastly = 1; Depends = 2; Eat together = 3); 2. Present
in the same room when husband is discussing with male members of the family; 3.
Participation in discussions with male members; 4. Husband helps wife in household
chores; 5. Accompany husband in social functions (Never = 1; Sometimes = 2; Always
= 3) and 6. Going to shopping alone or with husband (With female elders = 1; With
husband = 2; Go alone = 3). (Pooled minimum score per respondent is 6 and the
maximum is 18).
Extent of Restrictions Imposed on Women by Husbands to do Certain Activities
1. Going to shopping alone; 2. Going to film; 3. Spending money on
cosmetics, clothes, and so on; 4. Entertaining friends and guests; and 5. Financial
assistance to relatives (Does not allow = 1; Allows sometimes = 2: Always allows
= 3). (Pooled minimum score is 5 and the maximum is 15).
Extent of Women's Participation in Decision-making on Household Matters:
1. Spending money on food; 2. Spending money on clothes; 3. Health and medicine;
4. Inviting guests: 5. Decorating house; 6. Leisure time activities; 7. Spending
money on radio; television, and such capital goods; 8. Family size; and 9. Family
planning (Husband alone = 1; Husband more than wife = 2; Both husband and wife
= 3; Wife more than husband = 4; and Wife alone = 5). (Pooled minimum score is
9 and the maximum is 45).
2. Details of Techniques Used for Analysis:
(i) Hierarchical Analysis of Covariance:The hierarchical analysis of covariance is
a special form of covariance analysis where the covariate and factors are introduced
in a fixed order. The analysis shows for each variable, the contribution to the degree
of freedom and the sum of squares after controlling for variables listed earlier (but
not for variables listed afterwards). The ratio of the mean square for a variable to
the residual mean square, after introducing all the variables, gives an F-test for the
net effect of the variable after controlling for previous variable(s). The ratio of the
sum of squares for a variable to the total sum of squares provides a partial R2
measuring the degree of association between the variable and the dependent
variable, or the proportion of variance explained by the variable, after taking into
account the previous variables. Accumulation of partial R"s provide a multiple R2.
measuring the proportion of variance explained by all the variables considered up
to that stage (Sohail, 1981: Bhargava and Saxena, 1987).
(ii) Multiple Classification Analysis: The multiple classification analysis (MCA) is
a special case of analysis of variance with no interaction terms, and multiple
regression with dummy variables. Here the predictor variables can be nominal, and
a non-linear relationship between any predictor and the dependent variable is
permissible. The main advantage of the MCA technique is that it provides the grand
mean of the dependent variable as its constant term and a table of category of means
for each factor or predictor variable expressed as deviations from the grand mean.
Thus, they reflect the magnitude of the effect of each category of a predictors. These
category effects may be obtained in three different forms namely unadjusted,
adjusted for variations accounted for by the succeeding predictors, and adjusted for
variations in the predictors and for differences in the covariate as and when they are
appropriate (Bhargava and Saxena, 1987; Sathe and Murthy, 1987).

The Effect of Status of Women on Fertility 555
Audinarayana, N.
Socio-Cultural Dimensions at Marriage in Rural India,
New Delhi: Mittal Publications.
Female Age at Marriage and its Determinants in an Urban
Community of Tamil Nadu. Paper presented at the 17th
Annual Conference of IASP, Annamalai University, De-
cember, 16-19.
Audinarayana, N.
Status of Women and Fertility Behaviour in an Urban
and Thenmozhi, N.
Setting of Tamil Nadu. In A. Bose and M.K. Premi (Eds.),
Population Transition in South Asia, New Delhi: Mittal
Publications, 267-278.
Basu, A.M.
Culture and the Status of Women in North and South India.
In S.N. Singh (Ed.), Population Transition in India, Vol-
ume 2, New Delhi: B.R. Publishing Corporation, 233-244.
Bhargava, P.K.
Determinants of the Status of Women and Fertility in
and P.C. Saxena
Greater Bombay,. The Indian Journal of Social Work,
Dixon, R.B.
Rural Women at Work: Strategies for Development in
South Asia, Baltimore: The Johns Hopkins University
Dyson, T. and
On Kinship Structure, Female Autonomy and Demo-
Moore, M.
graphic Behaviour in India, Population and Development
Review, 9(1), 35-60.
Dubey, D.C. and
Status of Women and Fertility in India, New Delhi: Na-
Bardhan, A.
tional Institute of Family Planning.
Jejeebhoy, S.J.
Women's Status and Fertility: Successive Cross-cultural
Evidence from Tamil Nadu, India, Studies in Family
Planning 22(4), 217-230.
Mahadevan, K.
Women and Population Dynamics Perspectives from
Asian Countries, New Delhi: Sage Publications. 321 -341.
Mason, K.O.
The Status of Women: A Review of its Relationship to
Fertility and Mortality, Michigan: Population Studies
Centre, University of Michigan.
Gender and Demographic Change: What Do We Know?,
Belgium: IUSSP.
Mukherjee, B.N.
The Status of Married Women in Haryana, Tamil Nadu
and Meghalaya, Social Change, 1(4), 4-17.
Sathe, Y.S. and
Multiple Classification Analysis. In P.C. Saxena and P.P.
Murthy, B.N.
Talwar (Eds.), Recent Advances in the Techniques for
Demographic Analysis, Bombay: Himalaya Publishing
House, 339-358.
Sohail, M.
Differentials in Cumulative Fertility in Rural Bangladesh.
In Multivariate Analysis of World Fertility Survey Data
for Selected ESCAP Countries, Asian Population Studies
Series No. 49, Bangkok: United Nations.

556 N. Audinarayan
Sundar, R.
The Status of Women and Fertility: Some Field Results,
Monograph Scries No. 13, New Delhi: National Council
of Applied Economic Research.
Thenmozhi, N.
Status of Women and Fertility in an Indian Urban Setting.
Unpublished Ph.D. thesis, Coimbatore: Bharathiar Uni-
Visaria, L.
Regional Variations in Female Autonomy and Fertility
and Contraception in India. GIDR Working Paper No. 50,
Ahmedabad: Gujarat Institute of Development Research.
Vlassoff, C.
Progress and Stagnation: Changes in Fertility and
Women's Position in an Indian Village, Population Stud-
46(2), 195-212.