ECONOMIC ROLES OF WOMEN AND ITS

IMPACT ON CHILD HEALTH AND CARE:

A STUDY IN KERALA
A THESIS SUBMITTED TO THE UNIVERSITY OF MYSORE
THROUGH INSTITUTE OF DEVELOPMENT STUDIES,
UNIVERSITY OF MYSORE, MYSORE.

FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY IN

DEVELOPMENT STUDIES

BINITHA V THAMPI

UNDER THE SUPERVISION OF

PROFESSOR K. N. M. RAJU

POPULATION RESEARCH CENTRE

INSTITUTE FOR SOCIAL AND ECONOMIC CHANGE

BANGALORE - 560 072

JANUARY 2007

DECLARATION

I declare that the thesis entitled "ECONOMIC ROLES OF

WOMEN AND ITS IMPACT ON CHILD HEALTH AND
CARE: A STUDY IN KERALA" is the result of my own work
carried out at the Institute for Social and Economic Change,
Bangalore, and that it has not either wholly of in part, been submitted
for any other degree or diploma. Due acknowledgements have been
made wherever anything has been borrowed from other sources.

Date:

2=f 1t.f/
0
J~ 2001

Binitha V Thampi

Phone: 91 OSo-2321546S/5592/5519
FAX: 91-80-23217008

Website: www.isec.ac.in
Email: admn@isec.ac.in

INSTITUTE FOR SOCIAL AND ECONOMIC CHANGE

Nagarabhavi PO, BANGALORE-560 072
An all India Institute for inter-disciplinary research and training in the Social Sciences

CERTIFICATE
This is to certify that this thesis entitled, "ECONOMIC ROLES OF
WOMEN AND ITS IMPACT ON CHILD HEALTH AND
CARE: A STUDY IN KERALA" is a bonafide research work
carried out by Ms. Binitha V Thampi independently under my
supervision.

I also certify that this thesis has not been previously submitted for the
award of any degree or diploma or associateship to any other
University or Institution.

ktN~~
K. N. M. Raju
Professor (Rtd.)
Population Research Centre

ACKNOWLEDGEMENTS
My journey towards the completion of doctoral thesis is coming to an end. Let me
express my gratitude and feelings to those who have accompanied me in this journey.
First and foremost, my supervisor, Prof. K. N M. Raju, who always supported me and
without his encouragement this study would not have been completed. The kind of
academic freedom and moral support that he has extended to me is remarkable. I
thank him wholeheartedly for all his help and support.
It was certainly ISEC, which gave me an opportunity to join for the doctoral
programme and given me the fellowship without which I could not have carried out
this study. I am thankful to Prof. Govinda Rao, previous director of ISEC. Immense

was the help and support that he has given to me. I also acknowledge the help of Prof.
G.K. Kadekodi, previous director of ISEC.
My discussion with Dr. Janaki Nair has helped me immensely and I express my
gratitude to her. Dr. T. V. Sekbar and Prof. D. Rajasekbar have always extended their
support and encouraged me through the study period. I am duly acknowledging the
help and moral support of Prof. G.K. Karanth and Prof. R. S. Deshpande.
The help provided by various Ph.D coordinators such as Drs. K.V.Raju, G.S.Sasthri,
Usha Devi are fondly remembered. The present Ph.D coordinator Dr. Anand
Inbanathan deserves special mention. I am indebted to him for his timely support in
difficult situations. I also thank ISEC Registrar, administrative staff and library staff
especially, Mr. Kalyanappa, Mr. Venketesh, Ms. Leela, Mr. Rajanna, Mr. Suresh and
Mr. Rudresh. Special thanks are due to Mr. Krishnachandran, Ms. Margaret and Ms.
Shantha for all the help rendered during my study period. I also acknowledge the help
of Mr. K.S.Narayana in completing the study. I am thankful to all the members of
Population and Research Centre.
I happily remember Prof. P. Mohanan Pillai, Prof. T.M. Thomas Isaac and Dr. P. M.
Michael Tharakan from Centre for Development Studies, who have helped me at
various stages of my career and influenced my ideas and thoughts. I am grateful to all
of them.

In this occasion, I remember my friend J Devika from CDS. My academic

interactions with her has enriched and refined my thought process. Also, her affection
and encouragement helped me to complete this study. I am also grateful to my friends
Praveena and Beena from CDS for their support and constant encouragement. I thank
Ms. Sasikala for the editing work.
T. Kannan is a friend who accompanied me in my academic journey during last six
years. I enjoyed the academic debates and arguments with him and it often helped me
to refine my thoughts.
I am thankful to ISEC friends, Gagan, Emil,

Santhosh, Poulomi, Nisha, Lija,

Prasobh, Jeena, Jyothis, Amalendu, deepika, Prathibha, Tunga, Anand, sathyapriya

and Mahesh for the special company. Thanks are due to my colleagues as well as
friends at KILA, Mohanakumar, Madhu, Raju, Thresia, John M Itty and Raghavaiah
for their support. I also acknowledge the warmth and affection showered to me by my
friends Reshmi and Munawar.
My loving daughter Janu gave me the experience of the pleasure of mothering
without much pain. She has been a source of inspiration in my life. She is quiet
understanding and her patience helped me to sit and work on my thesis.
My family was highly helpful and without which the completion of this study would
not have been realized. My mother insisted me to continue my studies and happily
took care of my child when I joined for Ph.D. Since then, she has been encouraging
me to pursue and complete the same. I know that she will really be happy over the
completion of this work. I am sure that the affection and support of my sister Anitha
has helped me to complete the work. I remember Ajichettan, Chithra, Sreekumar,
Meenu and Devu in this occasion and their affection towards me. My husband, Shibu
has always been a source of encouragement and provided me sufficient space to make
my choices in life. Though I happily acknowledge the same, I am not venturing into
formally thanking him.
Last but not least, the women member of self help groups in Vellanad and Pallichal
whose help was remarkable and I acknowledge it with due importance. I am also
thankful to Mr. Krishna Kumar, who helped me in organising my filed work.

CONTENTS
Declaration
Certificate
Acknowledgements
List of tables............... ......... ............ ..................... ... ....... .. .. ....
v
List of figures............... ... ... ... ............ ........... ........................ .... vii
List of abbreviations ................................................................... viii
Chapter I............................................................................. ...
Introduction
1.1.
Introduction
1.2.
Larger context of the study
1.2.1. Women's work and its intrinsic value
1.2.2. The instrumental value of women's work
1.2.3. Caring labour and sharing of caring responsibilities
1.3.
Objectives of the study

1-8

Chapter 2... . ... .... ... .... . .. .. .. .... .. . . ... .... .

Conceptual Framework and Methodology
2.1.
Introduction
2.2.
Causal framework of factors determining child health and care
2.2.1. Explanation of the causal model
2.2.2 Review of the studies examining the impact of women's work
on child health and care
2.3.
Conceptual framework of the present study
2.4
Methodology
2.5.
The context of the study
2.6.
Chapter Scheme
2.7.
Data Sources

9 - 25

Chapter 3....... .... ........ .... .... .. . ..... ... .. ...

Understanding Women's W.ork
3.1.
Introduction
3.2.
Decentralized local governance, kudumbashree and economic
activities of women
3.2.1. The context of Kerala
3.2.2. Policy shifts in women's development
3.2.3. The Women component plan (WCP)
3.2.4. Kudumbashree self-help groups

27 - 48

I
3
3
5
7
8

9
9
10
II
16
18
21
22
25

27

29
30
31
33
34

3.3. Methodological issues in women's work

3.3.1. Undercounting of women's work

37
37

3.3.2. Multiplicity and intensity of work

3.3.3. Time use survey method to measure women's work
The nature and specificities of women's work in the study setting
3.4.
Summary
3.5.

39
41
43
48

Chapter 4 ............................................. 49 - 66
An analysis of the determinants of child health using
National Family Health Survey Data
4.1.
Introduction
49
4.2.
National family health survey data
50
4.2.1. Analysis of the determinants of child health using NFHS I data
57
4.2.2. Analysis of the determinants of child health using NFHS II data
60
4.3.
Discussion
64
4.4.
Summary
66
Chapter 5 .............. '" ......................... 67 - 83
The impact of women's work on child health and care -A quantitative and
qualitative exploration of primary data
5.1.
Introduction
67
5.2.
Methodological issues involved in the measurement of childcare time 68
5.2.1. Concept of active and passive care and time demands
69
5.2.2. Intensity and multiplicity of activities in the child caring
69
5.3.
Exploration of the association between maternal work and childcare
-~

5.4.

Analysis of factors determining maternal time spent

on childcare - primary Data
5.4.1 Discussion
5.5. Analysis of the impact of women's work on child healthPrimary data
5.5.1. Discussion
Qualitative data on work and childcare
5.6.
5.7. Summary

77
78
78
81
81
83

Chapter 6 , , . 85 - 96
Exploring the Relationship.between Women's Work and Female AutonomySecondary Data (NFHS II)
6.1.
6.2.
6.2.1
6.3.

Introduction
Factors determining autonomy - NFHS II data
Discussion
Summary

85
85
96
96

ii

Chapter 7 ......... ::................. 98 -114

Female Autonomy, Women's Work and Child Well-beingPrimary Data
7.1. Introduction
98
7.2. Female Autonomy and child well-being
99
7.2.1. Conceptualisation of female autonomy and women's empowennent 100
7.2.2. Operationalising the concept by employing appropriate measurement I02
7.2.3. Employing the concept of women's autonomy in child well-being
104
7.3. Women's autonomy and work status - Primary data
105
7.3.1. Results of the impact of dependent variables on
female autonomy indices
109
7.3.2. Discussion of the results of the impact on selected dependent
III
variables on female autonomy
7.4. Qualitative data on female autonomy
112
7.5.
Summary
114
Chapter 8 .................. ::.................................................. 116 - 124
Childcare arrangements -Role of Formal and Non-normal Institutions
8.1. Introduction
116
121
8.2. Anganwadi centres in Kera1a
8.3. Summary
124
Chapter 9 ................................................................................. 126 - 139
Impact of Maternal work on Paternal Care
9.1.
Introduction
126
9.2. Understanding caring labour
127
9.3.
Gender and caring labour
129
9.4. Paternal care as the distribution of caring labour
\30
9.5. Paternal care and maternal work - An Analysis
133
9.5.1 Discussion
136
9.6.
Qualitative data from focus group discussion
\37
9.7.
Summary
\39
Chapter 10 .................................................................................. 140 - 147
Working life and family dynamics -Reflection on Case Studies
10.1. Introduction
140
10.2. Changing dynamics of work
140
10.3. Intensity and multiplicity of work and childcare arrangement
145
10.4. Autonomy and decision-making
147
Chapter 11. ................................................................................ 149 -160
Conclusion and Policy Recommendations
11.1 Summary and conclusion
149
11.2. Future direction for public policy
160

iii

Annexure I .................................. 163-168

Profile of the study setting
Annexure II ....................................................... 169-178
Case studies
Bibliography........................................... 179-191

LIST OF TABLES

Table
no.s
3.1
3.2
3.3
4.1
4.2
4.3
4.4
4.5
4.6
4.7
5.1
5.2
5.3
5.4
5.5
5.6
5.7
5.8
5.9
6.1
6.2
6.3
6.4
6.5
7.1
7.2
7.3

Page Nos

Work participation rates in Kerala and India

Pattern of work
Time spent on work and childcare by typology of work
Prevalence of acute respiratory infection, fever and
diarrhea by background characteristics
Nutritional status of childrenby background characteristics
Anaemia among children
Details of the explanatory variables selected
Incidence of diarrhea among children on selected
explanatory variables - Odds ratios
Details of the explanatory variables selected
Incidence of diarrhea among children on selected
explanatory variables - Odds ratios
Percentage distribution of activities women perform
in the childcare regime
Time spent on childcare
Median duration and frequency of breast-feeding
Frequency of changing clothes and bathing
Percentage of morbidity incidence by work typology
Description of variable selected for the analysis of time spent on childcare
Regression results of determinants of time spent on childcare
Description of the explanatory variables selected for the
analysis of factors determining incidence of any illness among children
Incidence of any illness among children on selected
explanatory variables - Odds ratios
Household decision-making indicators - NFHS II
Women's autonomy by background characteristics - NFHS II
Women's experience with beatings or physical violence - NFHS II
Significance of dependent variables - Autonomy Index
Impact on dependent variables on female autonomy multiple classification analysis
Mean of autonomy indices by work typology
Significance of dependent variables on female autonomy Indices
Influence of dependent variables on female autonomy indicesmultiple classification analysis

31
43
47
52
54
56
58
59
62
63
71
72
73
73
75
77
77
79
80
86
89
91
94
95
108
109
110

8.1

8.2
8.4

8.5
9.1
9.2
9.3

Childcare provider when the mother works away from

home (0-4 age group) NFHS I '
Care provider for children in the 0-3 age group when
mothers work (NFHS II)
Primary care provider for 0-3 age group when
women work -primary data
Primary care provider for 3-7 age group when women
work - Primary data
Percentage distribution of paternal care by maternal work
Estimates of the ordered probit model- extent of paternal care
Marginal effect of independent variables on extent of paternal care

117
119
119

120
133
136
136

vi

LIST OF FIGURES

Figure no.s
2.1
5.1
6.1
8.1
9.1
9.2

Framework for examining the influence of female

work participation on child health and care
Percentage of minor accidents by work typology
Percentage of women participating in different household
decision and access to money
Percentage of children with mothers while working
Average daily time spent on childcare by fathers
Average daily time spent on'childcare by grandmothers
And others

Page no.s
II

74
88
118
132

132

vii

LIST OF ABBREVIATIONS
ADS
BPL
CBNP
CDS
FGD
ICDS
KSSP
KSWC
MCA
MOHFW
NFHS
NHG
OBC
PPC
PRis
SC
SHG
ST
UPAP
US
WCP

Area Development Society

Below Poverty Line
Community Based Nutrition Programme
Community Development Society
Focus Group Discussion
Integrated Child Development Services
Kerala Shasthra Sahithya Parishad
Kerala State Women's Commission
Multiple Classification Analysis
Ministry of Health and Family Welfare
National Family Health Survey
Neighbourhood Group
Other Backward Community
People's Planning Campaign
Panchayati Raj Institutions
Scheduled Castes
Self Help Group
Scheduled Tribe
Urban Poverty Alleviation Programme
United States of America
Women Component Plan

viii

Chapter 1

Introduction
1.1 Introduction

Academic concerns and policy efforts towards enhancing women's economic

participation during the sixties and seventies have paved the way for understanding
the impact of women's work on their families and vice versa. Many sociological
studies on the 'work -family' interface have been aimed at evaluating the effects of
maternal employment on childn:n and families as well as the effects of family factors
on work. (perry-Jenkins, Repetti & Crouter, 2000). Studies on work - family
spillover were begun to document the variables that contribute to work productivity
as well as to family functioning.
The work - family framework has helped to focus attention on the support that is
needed to minimize the negative spillover from work to the parent - child
relationship. The findings of studies on maternal employment and child health
outcomes bring a number of policy issues to the forefront because the domains of
work and family are embedded within the larger societal context. There is a wide
range of policy options that could have differential impacts on mothers, children and
employers etc.
When there was an increase in the percentage of women who entered the workforce,
researchers tended to design studies that measure the direct effects of maternal
employment on the experiences and development of their children. The underlying
assumption of these studies is thllt there must be some negative impact of their

mothers being in the workforce on children.

However, there are significant

inconsistencies in the findings of these studies in tenns of showing both positive and
negative impacts of maternal work on the development of children (Pleck 1985;
Thompson and Walker 199O).
There are studies in demography that specifically deal with the association between
women's work and child well-being.

A wide range of studies in the field of

demography focused on the positive influence of female education and employment

- the two major factors that embody the status of women - on demographic
outcomes. There is a pervasive consensus found in the literature over the strong
effect offemale education on fertility and child mortality reduction (Caldwell 1986;
Sen 2ooo). That analysis of the influence of female employment on child well-being
however shows mixed evidence, both positive and negative, laying more emphasis
on the social context in which the relationship takes place and being less definitive
about the direction of causality (Mamta Murthy 1995; Sivakami 2000; Busu and
Basu 1992).
The present study is an attempt to understand the association between women's work
and children's well-being in a,specific social setting. However, the interest of the

study is not limited only to the causal model that assesses the influence of maternal
work on child well-being. It also tries to locate the causal relationship of women's
work status on child well-being in a larger context, which is being infonned by
different scholarly concerns. In addition to examining the net impact of women's
work on child well-being using mUltiple data sets, it critically evaluates the
conceptual categories as well as the framework that assess the causal relationship.

Moreover, certain concerns pertaining to women's 'autonomy' and 'agency' as well

as the nature and characteristics of child care as 'caring labour' have also been
pursued from a feminist vantage point. Hence, the study is placed in a larger context
and the following is the description of the broader context of the present study.

1.2. Larger context of tbe study

As mentioned above, the central theme around which the present study has been
carried out is the causal relationship of women's work on child well-being.
However, the study focus is not limited to the causality of the formulated
relationship. It seeks to understand the larger context that has framed this particular
research question. It raises the following research questions: What is the relevance
and policy implication of the studies that analyse the impact of women's work on
child well-being? What are the problems in the existing conceptual categories that
are used in understanding such a relationship? How certain feminist notions, viz.
female agency, caring labour etc. give better insights to the present research problem
in terms of its political implications on women in particular and society in general?

In order to answer these questions, we need to discuss three distinct dimensions of

our central research theme that are primarily the vantage points with reference to
which the present study has beeir pursued.

1.2.1 Women's work and its intrinsic value

The recognition of women as a category in the development discourse has received

significant impetus since 1970s. Also, the various ramifications of women's socio.:conomic life have received more attention. Women'.s role in the economic

production has attracted research and policy focus (Boserup 1971; Beneria and
Roldan 1987; Dixon 1978). Women's work has also been acknowledged as one of
the major factors that enhance female autonomy. Again, it is expected that their work
outside their homes will bring about changes in the existing division of labour, which
is aligned along the gender lines. In addition, cash income as a major resource in the
hands of women could transform the gendered power relations within the domestic
domain. Hence, the intrinsic value of female employment has been emphasized and
it has been identified as one of the major driving forces behind women exercising
their agency. Along with gainful employment, I female education is also identified to
be a major significant factor in its own right. Amartya Sen (1999) emphasises the

recent shift in attention on women's 'agency' from the earlier 'welfares'

concentration on 'well-being'.

'No longer the passive recipients of welfare

enhancing help, women are increasingly seen, by men as well as women, as active
agents of change: the dynamic promoters of social transformations that can alter the
lives of both women and men' (p.189). The argument for the need to focus on
agency is precisely the role that such an agency can play in removing inequities,
which denies the well-being of women. This well-being is strongly influenced by
variables such as women's ability to earn an independent income, to find
employment outside the home, to have ownership rights, to be literate and thus be
educated participants in decisions within and outside the family. He argues that such
an agency has an intermediate role in enhancing women's well-being in many areas
of economic, political and social action, varying from rural credit and economic
activities, on the one hand, to political agitation and social debates, on the other (Sen

1------I

Occupations that provide direct income..

1999). Here, the understanding of female agency is not simply limited to women's
participation in gainful employment and acquiring educational skills, it reaches to
'public domain' where women assert their political will and engage in an array of
public political actions. The above description of female agency helps the present
study to develop an understanding of women's status. Better social status implies
that women can exercise their agency and acquire a certain degree of autonomy,
which positively influences both members of the family as well as the community at
large. The emphasis here is on the intrinsic value of work that provides women more
autonomy in the decision-making process.
1.2.2 The instrumental value of women's work
The instrumental value of female employment has received more attention in the
policy realm. It has been understood as one of the major factors contributing to the
reductions in fertility and infant and child mortality.
Increase in work participation of women outside the house tends to reduce the
desired number of children, as the opportunity cost of child-bearing and rearing is
high. Also, women who work outside home get better exposed to available options of
contraception. Since their economic contribution to the family is relatively high, their
decision-making power within the family on matters regarding fertility and
contraception is also high.

The literature suggests two mutually compensating effects of women's work

participation on child survival and care. The first one is the positive effect accruing

i from mother's income.

The mother's work nartidruotinn "'nh"n ......~ thl" family income,

which in tum has a positive impact on child nutrition and health.

A study by

Mencher (1988) on women in landless agricultural households reveals the different

ways in which men's and women's income is spent. The data presented strongly
indicate that a very high percentage of female income goes to household
maintenance, i.e. for food ani! health care of their children, whereas a lower
percentage of male income is comparably spent. The second is the negative price
effect derived from the opportunity cost or the shadow price of time the mother devotes
to childcare. Thus, the labour force participation of the mother can have an adverse
impact on child health as the child may not get full attention from its mother and may
even have to forego the benefits of breast-feeding. This probably happens in those
families where, because of poverty, the mother must participate in the labour force soon
after delivery. Hence, whether the net effect is positive or negative would be context
dependent.
The above explained causal pathways of influence of work participation on fertility and
child mortality reduction sees primarily the extrinsic value of work. This is an
important approach adopted in demographic studies and the results of these studies are
extensively used for policy formulations as well.

It is expected that women's employment may translate into greater control by them
over the expending of resources, increased exposure and access to relevant
information about child-bearing and rearing practices and an enhanced ability to
manipulate and engage the world outside the home to meet the nutritive. medical and
survival needs of children better (Sivakami, 2000). Hence, in order to capture these
bffects of maternal work on child health and well-being, the concept of 'Female

Autonomy' has been considered as an intermediate variable in the causal model of

women's work and childcare and health. Nevertheless, it should be noted that the
instrumental value of female autonomy in enhancing child health that women acquire
through their work outside the home is focused upon here.

1.2.3 Caring labour and sharing of caring responsibilities

Since the main responsibility of biological reproduction lies with women, the rearing
and caring of children are traditionally considered its natural extension.

It is

expected that women's work outside home might bring about changes in the existing
division of caring labour on gender lines within the household. The share of men in
the childcare regime will naturally increase as maternal work participation increases,
which in tum alters the existing gender division of caring labour significantly. This
helps to reorganize the social life of both men and women.

In the context of advocating female employment outside home as a strategy for

improving their socio-economic status, the impact of such employment on their
primary role as childcare providers and the question of the compatibility between the
two assume relevance in the policy realm.

The policy prescription towards

promoting home based income-generating activities for women, providing creche

facilities near the workplace etc. are efforts to reduce this postulated incompatibility.
However, it has also been noticed that other institutional arrangements for childcare
would have significant bearing, on the outcome. The family remains as the major
social institution for childcare in less industrialized societal settings. Also, there are
,public state-owned institutions like anganwadies, which run pre-schools. Private

childcare centres are also available that are more prominent in urban areas and
working mothers belongs to middle class and upper middle class often utilize
privately run creche facilities.
It is against this backdrop of the larger context with three identified dimensions
pertaining to women's work and childcare which significantly inform the research
concerns being considered in

the present study, that the objectives of the study are

formulated as follows.
1.3 Objectives of tbe study

I.

To understand women's work and its specificities in the study context

2.

To examine the differentials in child health status in relation to the work

characteristics of women at the macro and micro level.

3.

To analyse the causal mechanisms in the association between women's work

and child health care.

4.

To understand female autonomy in relation to their work status and child wellbeing.

S.

To appraise the different institutional arrangements existing for childcare

6.

To study the nature and determinants ofpatemal care and the distribution of the
caring labour within the family.

Cbapter2

Conceptual Framework and Methodology

2.1. Introduction
In the broader context of the study that has been elaborated in the previous chapter
and the objectives that have been formulated, the conceptual framework and
methodology of the present study has been developed. It will be discussed in detail
after explaining the causal model that examines the relationship between women's
work and child health and care and the literature review related to it. Since there are
studies arguing that maternal work affects adversely the child health outcomes, the
present study uses the causal model to examine the impact of women's work on child
health and care at the macro and micro level.
2.2. Causal framework of factors determining child health and care
Since the central theme of the present study is to understand the relationship between
women's work and child health and care, it is important to understand the causal
framework that has been employed in the existing studies and the results that arise
from them.

The following is the conceptual frame of the formulated causal

relationship between women's work and child health and care and a review of the
existing studies that examine this relationship.

2.2.1. Explanation of tbe causal model

The above conceptual framework has been used by various studies examining the
causal relationship between maternal employment and child health and care. It is
expected that the set of background variables such as caste, class, family structure,
educational composition of the members of the household etc. influence the type of
work women do, as well the nature of childcare support available when women work
outside the home. Also,

the~

background variables determine the household

allocation of income on childcare. The autonomy that women acquire from their
work outside the home enhances the household resource allocation in favour of
children and it helps them to avail better preventive and curative care as well as to
purchase more nutritive food for children, all of which positively influences child
health. Again, the autonomy that women acquire might facilitate them to access
relevant information on child-rearing practices and access better nutrition and
medical care for children. All these have positive impact on child health. Sometimes
work participation may also

ha~e

an adverse effect on child health and care owing to

the reduced amount of time that working mothers can devote to their children.
Institutional mechanisms available for childcare have a decisive influence on
compensating for maternal absence. Hence, the net effect of maternal work on
childcare would be context dependent to a great extent. The following is a review of
selected literature that specifically looks at the impact of women's work outside
home on child health and care.

10

Fig. 1.1 : Causal framework for enmining the influence oUemale work participation
on child health

,---.

-.

Background
factors

r--

f--

Household
income

/
Female
autoJ1omy

-.

-.
---+

Female
work
Participation

Institutional
and familial ---+
support

~.

Allocation of
resources for
health

Nutrition

CHILD
HEALTH
& CARE
Awareness of
health care
resource
Access to health
care

t
Preventive
and
Curative
Health
care

Time spent on child care

2.2.2 Review of tbe studies eumining tbe impact of women's work on cbild
bealtb and care
In India, many studies have analysed the impact of women's work outside home on the

survival chances as well as health of their children and the following are the findings of
such studies.
Using Census data of 1981, Yearbook of Ministry of Health and Family Planning 1987
and Registrar General ofIndia 1988, Basu and Basu (1991) came to the conclusion thal
~omen's employment, in spite of other benefits, has a crucial adverse consequence on

II

child survival. By examining v,arious intennediate relationships, they conclude that

shortage of time is one of the main reasons for this negative relation between maternal

employment and child survival. However, this study does not have adequate evidence
on the distribution of women's

WOlX

by kinds of employment, Again, since like non-

working women, working women are also heterogeneous in terms of their

characteristics, it is possible that they share a common disadvantage which predisposes
them to greater child loss, may be low income, The authors argue that while employed
women may be forced to work because of their greater poverty, the additional income
from their working has to reduce their poverty, Here, it is important to see whether this
income is supplementing or substituting the male income of the household and the nonavailability of such information from macro data like the Census makes the analysis
inconclusive,

The study by Sivakami (1997) examined the linkage between mother's work and child
health, The data are from a survey of 75 working and 75 non-working women in a
village in Tamil Nadu, India. The results show that the working women spend on an
average 2 hours less than non-working women on childcare, The duration of breastfeeding is also shorter among working women. Morbidity rates were higher for the
children of working women than for the children of non-working women. Nevertheless,
the break-up of time spent on different activities of childcare shows that the major
difference in the time spent on childcare among working and non-working mothers is
that spent in playing with their children (i.e. 1.8 Hrs. out of 2 Hrs). Also, though the
average duration of breastfeeding is shorter among working women, the difference is
rot high enough to make any significant adverse impact on childcare (12 months and

12

14 months among working and non-working mothers respectively). What appears more
important is that in the sample taken, most of the working women are wage labourers

and all indicators of their socilHlCOnomic background show that they lag far behind
their non-working counterparts in most of the selected aspects. Further, analysis shows
that work status turned out to be a significant variable in explaining one of the child
health indicators, i.e. the provision of special food for children, which indicates the
sheer economic pressure within the household.

A study done by Gillepsie and McNeil (1992) on the dynamics of the food system in
South Indian village society and its relation to agriculture, social organisation, physical
work, nutrition, health and survival has analysed the impact of maternal work on child
health and nutrition.

It was found that the net effect of maternal labour force

participation and time allocated to activities of household work are incompatible with
childcare using correlation analysis. They argued that the adverse effect of high
maternal labour participation on child nutritional status is likely to have impact on both
quantitative and qualitative aspects of childcare and feeding.

Infants tended to be

'opportunity-fed' as opposed to 'demand-fed', i.e. they were fed whenever the mother

had a chance. Even though the correlation results shows a high degree of association
'.

between maternal labour force participation and child malnutrition, it did not prove to
be the cause.

Using aggregate data from the Registrar General of India's survey of infant and child
mortality, Beenstalk and Sturdy (1990) concluded that, controlling for other socioeconomic factors, the relative probability of infants dying was 27percentage higher for

~orking mothers than for non-working mothers.

13

While analysing the impact of mothers' employment on infant and child mortality in
India, using NFHS data 1992, Kishore and Parasuraman (1998) found that mothers who
are employed have \0 percent higher infant mortality rate and 36 percent higher child
mortality rate than mothers who are not employed.

Hence, they arrived at the

conclusion that employment of women outside the home for cash, perhaps the most
empowering form of employment, is associated with elevated risk of infant and child
mortality.
Unfortunately, the above two studies have ignored the selectivity in female
employment. In India, women who are involved in market work are more likely to be
poor, landless and belonging to Ute backward castes. Some of the studies in the Indian
context have recognised this dimension of women's work and incorporated it in their
analysis.

Krishnaj i (1995) has analysed the impact of female work status on child survival using
1981 Census data and found that in all the states of India, working mothers experience a
greater child loss than non-working mothers. He arrived at the conclusion that in India,
work among women can be closely identified with poverty and this identification of
poverty of a lasting kind with working mothers, results in poor nutritional status and
health care among their children:'

Using the Census data of 1981, a district-level analysis of the determinants of mortality,
fertility and gender bias in India has been made by Murthy, Guio and Dreze (1995).
They realised that in estimating the effect of female labour force participation on child
r0rtality and other demographic outcomes, it is important to control for the incidence

14

of poverty. Indeed, the female labour force participation in India is often a reflection of
the economic hardship, and the failure to control for this factor may lead, for instance,

to a spurious positive relationshlJ> between female labour force participation and child
mortality, while implicitly reflecting the positive association between poverty and child
mortality.

The analysis, after controlling for this variable, showed higher female

participation is associated with lower child mortality.

Desai and Jain (1994) have examined the role of gender inequality within the context of
the political economy of rural production relations in shaping intra-family dynamics
associated with women's employment and particular, the relationship between maternal
employment and child welfare in the context of gender inequality in the labour market
and the poverty and lack of access to infrastructure in ruraJ South India They argue
that in many developing countries, women's domestic burden may pose a great
impediment to their participation in those economic activities that may yield higher
incomes. Consistent with Basus' findings on child mortality, it was found that in
villages in Kamataka, children whose mothers are wage workers are less likely to be
immunised and more likely to suffer from poor nutrition than children whose mothers
are family workers or housewives. On the other hand, the difference in health and
nutritional status between children whose mothers work on the family farm and whose
mothers are housewives is relatively small. However, this result has not led them to
conclude that maternal participation in wage work causes poor health outcomes for
their children. Rather, they have seen it as an outcome of the socio-economic context in
which women work. Women's economic activities are strongly correlated with family
ftcome and socio-economic status. Wage workers and petty traders are more likely to

15

belong to SCs and STs, to have less education, to have lower levels of fiunily income
and consumption. Therefore, the apparent negative correlation between mothers wage
work and child health is likely to be caused by the family'S socio-economic
circumstances and may have little to do with maternal work status. Again, based on
actual time-use patterns, it was found that the mother's time in specific childcare
activities such as feeding or bathing children does not differ much by work status, but
the mother's leisure time gets affected.

The above review suggests that there are many socio-economic, cultural and
demographic factors, which influence the association between women's work status
and child care and survival. In this context, the present study intends to focus on this
complex relationship. However, the present study does not limit itself to within the
above mentioned focus of assessing the causal structure. In addition, it seeks to
understand the working lives of women, their agency and autonomy, patterns of
childcare and sharing of the care among members of the family etc. that would help us
to understand the central research theme in a better manner. Hence, in order to
understand these dimensions of women's work and childcare, the following conceptual
framework has been adopted.

2.3. Conceptual framework of the present study

The study focuses on the relationship between women's work and child health and care.
Studies in the area of demography have examined the various pathways of influence of
women's work outside home on child well-being in different settings and come out
with varying results. The present study adopts the same causal framework and carries

16

out similar analysis. Nevertheless, it also attempts to make a careful scrutiny of the

causal framewod< that calls for broader understanding of the formulated relationship
under the study. It leads us to understand various conceptual notions related to women's
wod< and child well-being.
First, the question of the right to wod< and equal employment opportunity that has been
articulated by the feminist scholarship, acknowledges the intrinsic value of wod<. It is
expected that women's wod< outside home would alter the existing labour division on
gendered lines and lead to negotiation over gender norms within the domestic sphere. It
is also anticipated that wod< provides women more control over household resources
and helps them to make autonomous decisions. Hence, an attempt has been made to
understand women's wod< and its specificities in the study setting, and the nature and
degree of autonomy that women derive from their working lives outside the home.

Second, the instrumental value of wod< in terms of reducing fertility and enhancing
child well-being has received considerable attention in the policy discourse. Since the
focus is on the instrumentality of work, it measurers the value in terms of its power to
achieve the desired demographic outcomes. The direction of causality and the net
effect of the same are important in this approach.

The framewod< uses female

autonomy as an intermediate variable in the relationship between women's wod< and

child well-being. It is expected that the wod< outside home will enhance the survival
chances and health status of children through the increased capacity of mothers to meet
timely childcare requirements. The present study also makes an assessment of the
instrumental value of wod< in terms of its impact on identified child health outcomes,
.bing both primary and secondary sources of information.
17

Both the above postulations use the concept of female autonomy. Female autonomy
has been viewed as an outcome of the exercise of female agency in the fonner, whereas
it is conceptualised as the ability or capability to allocate more economic resources to
the family and provide modem childcare facilities in the latter. Therefore, it calls for
critical evaluation of the concept of autonomy, and thus an analysis of the relationship
of female autonomy with women's work status and child health has been carried out in
the study.
Third, it addresses the question of the tension between women's primary role of
mothering and the paid work. Caring labour comes under the broad category of social
reproduction. As mentioned before, it is expected that women's work outside home
would result in altering of the favourable sharing of caring labour within the family.
Using data collected from field surveys, an examination of the sharing of childcare by
the spouse has been done. Also, other institutional arrangements for childcare such as
public care facilities have been analysed.
2.4 Methodology

The above discussed conceptual framework necessitates an interdisciplinary approach

to the present study. Though the central theme of the present study is to examine the

relationship between women's work, and child health and care, which is being
articulated within the logical framework of demographic research; it borrows from
feminist notions, particularly while discussing female autonomy and caring labour. The
feminist schohuship has grown ~ a significant body of knowledge through its critical

18

encounters with mainstream social science disciplines and hence is interdisciplinary in

its approach.
Scholars both within demography and outside have raised concern over the lower
accommodativeness of gender both as a conceptual and analytical category in the
discipline of demography (Presser 1997; Raily 1999; Stancey and Thome 1985). The
primary impact of the wom~n's movement on demographic research in the
developing countries in the 70s and 80s was its focus on the relationship between the
"status of women" and fertility, at the micro level, using quantitative measures of
status, primarily women's education and employment, and generally keeping within
the analytical perspective offertility as a dependent variable.

In her profound article published in the journal, Population and Development Review,
Presser (1997) makes observations on the traverse of academic pursuit in the
demographic discipline with its overwhelming influence of international policy-making
bodies and its funding organisations. Even in the context of recent changes that led to
the so-called paradigmatic shift in the field of popUlation, she notes that academic
research has not been equally receptive to incorporating gender.

Gender issues have become central in the pol icy arena, as seen in
the ICPD while remaining marginal in demographic research. The
political centrality of gender issues, amounting to a paradigm shift,
is evident in the Program Of Action of 1994 ICPD conference, held
in Cairo, with its extensive attention to women's welfare and issues
of equity and equality vis-a-vis men. Women's movement around
the world operates mostly through the Non Governmental
Organizations (NGOs) that pursue gender politics have had a
decisive role in the realm of policy making. There have been
debates and serious concerns placed by scholars both in the
academia as well as in the policy realm over the newly acclaimed

19

space of feminist activism within this major international policy

formulation forum on population and demographic forums.
Although some demographers have actively been promoting gender
issues from a gender perspective- specifically, research that
considers gender relations and the well-being of women relative to
men as both as a cause and consequence of demographic processes
and that aims to promote gender equality-such analysis have been
few and marginal to the field.
(Harriet B Presser, 1997)
Attention to gender in demography has been primarily through an approach that can
be described as feminist empiricism, one that primarily involves a focus on including
women in the existing models. It is invariably an individual level approach to gender
that focuses on individuals and collects data at the individual level, as demographic
research often does, which limits understanding of how gender influences social
behaviour. Stancey and Thome have noted that this approach essentially reduces
social life to a series of measurable variables, diminishes the sense of the whole that
is crucial to the theoretical understanding of social, including gender, relationships
(Stancey and Thome 1985).
The major argument against using quantitative methods in understanding gender is
the argument against positivism in general. The three major issues regarding
positivism in our context are: (I) positivism's adherence to the notion that physical
science is the best model for social research; (2) notions of universal laws which
assume that regular relations between variables are held to obtain across all
circumstances, and (3) a belief in objectivity, that research is or can be conducted
divorced from the politics and social relationships of power, class, gender and race of

20

the society in which it takes place. These notions have been critiqued from many
fields and have specifically been strongly attacked by feminist theory (Riley, 1999).
The methodological stance adopted in the present study is that while acknowledging
the value of quantitative data and methods in documenting certain inequalities, it is
important to recognize gender as a social construct that goes beyond an
understanding of inequalities at individual level variables and its all pervasiveness as
one of the organizing principles of all societies. Hence, the complementarity of
qualitative and quantitative methods is acknowledged and adopted. This is not only
so in the case of understanding gender but also holds good in the case of
understanding various social aspects of women's life, such as work, childcare etc.
The specific trajectory of inquiry adopted in the present study is to quantify the
formulated relationship and then to seek qualitative information to either supplement
or reconceptualise the relationship under study. An attempt is also made to locate
instances where gender figures significantly in the observed patterns of relationships
using both quantitative and qual,itative data.
2.S. The Context of the Study
The study is contextualised in the state of Kerala. The Kerala context is characterised
by low levels of female work participation coupled with better demographic features
in terms of child health compared to the rest of the Indian states. The infrastructure
facilities in the health sector are relatively better off in Kerala in comparison to the
other states, with better transportation facilities. The level of poverty too is in
variance with the rest of India, indicating a higher standard of living, lower fertility

21

and fewer children etc. which characterizes the general context. This might yield a
different result in terms of the impact of women's work on child well-being.
The current development debates on Kerala have pointed out the invisibility of
women in the production process and call for policies to enhance women's work
participation. State initiatives are also being made to enhance women's income
generation through promoting self-help activities among the poorer sections and to
set

apart a special plan component of 10 percent of total plan funds exclusively for

the development of women, sJX:Cifically focusing on their economic well-being. It is

in this context that a study analysing the impact of work on child health and care
assumes significant.

2.6. Chapter Scheme

The chapter scheme of the present study is developed based on the above discussed
conceptual framework and a brief description of the same is provided below.

An introduction to the study topic, its relevance, and the formulated objectives are

provided in chapter one. The causal framework to examine the relationship between
maternal work and child health and care is explained, and a review of relevant
literature examining this relationship is included in the second chapter. In addition,
the conceptual approach of the present study, the methodology adopted, and the data
sources are discussed.

In the third chapter, a detailed account of the study setting is provided, that takes into
~unt the recent economic restructuring of women's working lives around micro-

22

finance activities which has sis.nificant bearing on their working lives and the time
allocation for childcare. The conceptual issues in understanding women's work
within and outside the domestic sphere and the specificities of such work have been
discussed. With the help of a time use survey, an appropriate work typology has also
been developed and explained in this chapter.
Chapter four provides a multivariate regression analysis of the causal relationship of
explanatory factors including women's work on a selected child health outcome, i.e.
morbidity using NFHS data, Rounds I and II. It uses the causal framework that is
explained in chapter two.
Using both primary quantitative data and qualitative information collected through
focus group discussions, an exploration of childcare regime and various activities
involved in the same are discussed in chapter five. The impact of women's work on
time spent for childcare as well as child morbidity has been assessed in this chapter
using multivariate regression analysis. The qualitative data have been used to
interpret and supplement the results obtained in the quantitative analysis.
Chapter six includes an analy~is of factors determining female autonomy using
NFHS II data. Since female autonomy is being considered as an intermediate
variable in the relationship between women's work and child health outcome, its
nature and determinants have been analysed in detail. The impact of female
autonomy on child health has also been discussed in this chapter.
Based on the results obtained in chapter six, pertaining to the characteristics of the
female autonomy as collected in NFHS II and its relationship with child health

23

outcome, a critical evaluation of the concept of women's autonomy, which is widely

employed in demographic stu<;lies as an intennediate variable in the association
between women's work, and child well-being is carried out and incorporated in the
chapter seven.
The existing institutional arrangements for childcare (both public and private) would
have significant bearing on detennining the impact of maternal work on childcare in
any given context. Hence, an evaluation of the same has been carried out in chapter
eight. It uses both primary and secondary sources of data.
It is expected that women's participation in market production results in enhancing
their autonomy and a negotiation over gender roles within the family. In this context,
the feminist interpretation of caring labour and its fairer distribution between sexes
within the family have been discussed in chapter nine. The extent of paternal care
and an analysis of the impact of maternal work on paternal care have been included
in this chapter. Primary field data and qualitative data have been used in this chapter.

In order to further our understanding on the intersection of gender, work and family,
selected case studies have been conducted. A reflection on such case studies is
provided in the tenth chapter.
Chapter eleven finally includes the conclusion of the study and the relevant policy
directions that emerge out of the findings.

24

2.7. Data Sources

The study employs multiple data sets that include both quantitative and qualitative
infonnation. Quantitative data include both primary and secondary source of
infonnation collected from the study field. The secondary data include those from two
rounds of the National Family Health Survey conducted during 1992-93 and 1998-99.
A primary survey using an interview schedule has been carried out between 2002
September and 2003 January.

While analysing quantitative primary data, certain

research themes have been identified that require further field exploration and a
qualitative round has been conducted on these selected themes. Eight Focus Group
discussions were carried out with these selected themes and ten case studies were also
done for in-depth analysis.
In order to conduct the primary survey, two village panchayats have been selected
from the Thiruvananthapuram district of Kerala State. They are Vellanad and
PaJlichal

panchayats

30

and

18

kilometers

away

respectively

from

Thiruvananthapuram city. It is argued that much of the adverse consequences of

maternaJ work on child well-being are induced by the effect of poverty rather than
the work per se. Hence, in order to understand the association in a better way, it is
better to conduct the study amo~g the poor rather than apply statistical controls over
observed social realities.

Households have been selected from the BPL (Below

Poverty Line) list of the village panchayats and identified currently married women
having at least one child below the age of seven. There were 842 such women in this

2S

category from both the panchayats and a random selection of half of them, i.e. 421
was taken as the study sample.
The analyses using NFHS data. are not limited only to poor women. Though it
provides information on the standard of living index categorized as low, medium and
high, due to small number of children in each category, disaggregation has not been
done while analysing the impact of work status on child morbidity incidence.
However, the living standard has been considered as an explanatory variable.

26

Chapter 3

Understanding Women's Work

3.1. Introduction

"

One of the major objectives of the present study is to understand women's work
outside home in terms of its specificities. Such specificities in the domain of
women's work have been documented by various studies to a certain extent in the
context of developing countries (Baneria 1982; Bruce and Dwyer 1988; Donahoe
1999). Since one of the important themes of the present study is to understand the
ways in which women's work influences the caring responsibilities of women and
how far it transforms their own space within the family and community by giving
them an individual agency, it is important to understand 'women's working lives'
conceptually, as well as within its specific dimensions in the study setting. Therefore,
it is attempted to explore various aspects of women's work conceptually and the
same has been analysed in the context of the study field.
The context of the study field, the state of Kerala has undergone significant changes
in the direction of redefining the economic space of poor women during the recent
past. The decentralization process carried out in the State has incorporated a special
plan component, Women Component Plan (WCP) and organized women into selfhelp groups under a statewide programme called "kudumbashree".

In order to

understand the changing nature of women's work in the study setting, these changes
have been traced in the present chapter.

27

The conceptual issues involved in understanding women's work such as noting the
undercounting and undervaluing of women's work, and the specificities of women's
work viz. the mUltiplicity of activities and the intensity of work time, have been
discussed. To address these issues effectively, an appropriate methodology of data
collection is employed.
Quiet often, studies that analyse the causal relationship between women's work and
child well-being use the conventional understanding of work. They use the concept
of labour force participation as employed by the official statistical collectors of
information on women's work. As a result, they adopt a dichotomous categorization
of women's work, as in the labour force and outside the labour force. The
assumption made while using this categorization is that mothers who are in the
labour force have considerably less amount of time at their disposal for childcare,
which might have significant' adverse consequences on the well-being of their
children. However, all working mothers or non-working mothers may not be
categorically homogeneous in terms of the time at their disposal for childcare. The
hypothesis that is proposed here is that there are significant variations among
working and non-working women, in terms of the time input for childcare, that
results in varied impact on child well-being. Hence, it has been attempted in this
chapter to develop a new typology of women's work by understanding the varied
nature of women's work. In order to capture the specificities of women's work and
to develop a typology of the smire, a time use survey has been employed. In addition,
a focus group discussion is also carried out to understand the specific nature of
WlX11en's work.

28

3.2. Decentralized local governance, Kudumbashree and economic activities of

women
There has been significant restructuring of the economic activities of poor women
around micro-credit activities in Kerala during the last one decade. Unlike in other
parts of India, the micro-credit intervention in Kerala is structurally different and is
marked by the specific intervention of the State in both organizing women into selfhelp groups and organically linking them up with local governing structures. This in
tum has made substantial changes in the organising pattern of working lives of poor
women that calls for a new understanding of women's work. It is imperative now to
trace the trajectory of the micro-credit intervention framed within the structure of
local governance in Kerala in order to understand its specific stake in the
restructuring process. The following is a description of the same.
The Peoples Planning Campaign (pPC) for decentralised planning during the period
of the Ninth Plan (l997-2002),.carried out in the state of Kerala has made overt
efforts to address the issue of women's economic invisibility in the development
process. The increasing economic marginalisation of women in the context of the
high rates of unemployment that is being exhibited by the state, along with the
changed understanding of the concept of women's development in recent years, have
detennined the policy approach and institutional structures envisaged for this new
effort.
The state of Kerala has received wide attention for the paradoxical nature of its
development pattern, characteriSed by high achievements in social sectors on a weak
economic base. However, even though the achievements in social sectors are

29

remarkable, the female work participation rates in Kerala are among the lowest when
compared to that of other states. Hence, one of the development crises that Kerala
faces is that of increasing economic marginalisation of women in the development
process. Before looking at specific efforts to reduce this economic marginalisation,
it is important to discuss here the local and larger contexts that shaped such an effort.
3.2.1. The context of Keral.

The invisibility of women in the public domain of Kerala, despite the commendable
improvement in their educational and health status, is a paradox that has drawn
considerable attention during recent years. The economic marginalisation of women
in the development process can be seen from the table given below. Not only is the
work participation rate of women in Kerala significantly lower than the all-India
figure, it also tends to decline further. While the female work participation rate in
India has increased from 22.73 to 25.70 per cent between 1991 and 2001, in Kerala it
has continued to decline from

I~,85

to 15.30.

An important reason for the decline in the female work participation rate of women
in Kerala is the severe unemployment rate in the State. The unemployment rate in
Kerala is three times larger than that in the rest of India. Further, the incidence of
unemployment among females in the state is higher than that among males by two
times in the rural areas and three times in the urban areas (Prakash 1988;Ommen
1992). Such is the severity of unemployment that women are discouraged from
actively seeking employment and as a consequence, they drop out of the workforce
(Mathew 1995).

30

Table 3.1: Work particIPation rat es .In

Kerala
Year
Female
Male
19.70
65.66
1961
16.61
44.89
1981
15.85
47.58
1991
15.30
50.40
2001

K eraIa andId'
n la
Total
42.68
30.53
31.43
32.62

Male
51.5
52.62
51.56
51.90

India
Female
27.9
19.67
22.73
25.70

Total
42.5
36.70
37.68
38.57

Source: Economic Review, Government of Kerala, 2003

3.2.2. Policy shifts in women's development

The growing evidence on the increasing economic marginalisation of women in
Third World countries and the demand from women's groups for 'gender justice' in
the development process have invoked certain policy responses. In India too, the
policy approach and the institutional framework towards women's development have
undergone considerable changes.

During the fifties, it was a welfare-oriented

approach that emphasised family centered programmes. A marked shift took place
in the seventies with the Women in Development (WID) approach that viewed
women as active participants in the development process through access to
employment.

However, the concept of integrating women into the development

process has often ended up in simply adding them into the existing programmes. It
did not reflect upon the gendered structural constraints that systematically
discriminate against women even in situations of equal opportunity.

A better

conceptualisation in terms of gender as a social construct and unequal power

relations between men and women, was the characteristic feature of the Gender and
Development (GAD) approach. Nevertheless, it has not resulted into the planning
practices. Gender-sensitive planning is not an easy task, mainly due to the fact that
'.

many of the norms and practices we follow appear to be natural and justifiable, and
31

hence call for careful gender analysis. Also, it requires concerted efforts at the level
of policy formulation and at all levels of implementation. The recent approach of
empowerment is a useful tool that emphasises on empowering women, and put in
central place the individual agency of women in the process of their empowerment.
It stresses on the role of women's organisations in meeting their needs, both practical
and strategic, in recognition of the limitation of the top-down government approach
(Eapen el al 200 I)

These recent shifts in terms of the conceptual understanding of women's

development have paved the way for new policy approaches that emphasise both
economic and political empowerment of women. The 73rd constitutional amendment
for 113rd reservation to women in the Panchayati Raj institutions was a conscious
effort towards politically empowering women. Economic independence is viewed
primarily as a major component of the empowerment process. However, the
recognition of the economic independence of women as an indispensable factor of
their empowerment process has been realised at the programme level as part of the
anti-poverty efforts. As a resul~ the setting up of self-help groups of women has
evolved as an alternate institutional mechanism to address women's poverty and
promote economic empowerment. The self-help groups are considered to have an
advantage over bureaucratic institutional set-ups as being better institutional
arrangements with more democratic features. Hence, it has been decided to
implement all major centrally sponsored programmes through the self-help groups of
women. To match with the conceptual, policy and institutional shifts in addressing

32

women's development, planni~'g tool also need to be reformulated and gender

planning has became an accepted tool.
3.2.3. The Women Component Plan (WCP)

In Kerala, the emergence of the new generation of Panchayati Raj Institutions (PRls),
with one-third representation of women, has been accompanied by a massive campaign
to decentralise the planning process in the State. Nearly 3S to 40 per cent of the annual
plan funds of the state government are being set apart for projects and plans drawn up
by the local government institutions. These institutions have opened up space for local
initiatives to address the development crisis that the state is undergoing.
It is in the above discussed local and larger context of gender planning, that two

major steps were taken to enhance the income-generating activities of women in

Kerala. The first is the introduction of a new plan component, earmarking ten
percent of plan funds for women beneficiaries, known as the Women Component
Plan (WCP), considered as a unique feature of the KeraJa's decentralized planning.
Another major step was the initiation and promotion of self-help groups of women.
From the experience of the Bangladesh Grameen Bank and other similar experiments
elsewhere, the potential of self-help groups is acknowledged as an institutional
mechanism to promote income-generation activities of women by involving them in
thrift and credit activities and micro-enterprises. Besides, the potential of self-help
groups as micro level institutions through which the principles of democratic
participation in the local level decision-making process could be promoted was also
realised. As a result, efforts were made to link the micro-credit programme with the

33

structural changes in the rural areas, the emerging PRIs and participatory local level
planning. Hence, self-help groups came to be accepted as an ideal institutional
mechanism for the implementation and management of many of the development
projects created under the WCP (Issac and Franke, 200 I)
In the first year of the planning (\ 997-98), it was suggested to the local governments
that they earmark 10 percent of their plan funds to women beneficiaries. However,
since only 4 percent was actually allocated in the first year, the earmarking of 10
percent was made mandatory from the second year onwards. The point that needs to
be stressed in this context is thlJ! WCP has opened up a space for local initiatives to
generate employment opportunities for women belonging to the poor economic
strata. The nature and characteristic of labour supply vary considerably across
pachayats.

Hence, employment opportunities could be generated, taking into

consideration specific characteristics such as levels of education and other skills.

Realising the significance of such basic information for the planning of development
projects particularly for women, it was suggested to the local governments that they
undertake a 'Status of Women Study' exploring all the possible dimensions of

..

women's status during the initial years of the Peoples Planning Campaign. In the
case of planning projects under WCP for creating employment opportunities for
women, this information would be a valuable source.
3.2.4. Kudumbashree self-help groups
Another institution that has evolved at the local level along with the decentralized
plan!ling process in Kerala is the self-help groups of women.

Self-help groups

34

(SHGs) of women were accepted as democratic and women-friendly bodies for the
economic empowerment ofw0'!len. The Kudumbashree programme of the State, for
poverty eradication by organising women's self-help groups, has brought them under
the local governments and integrated them with the functioning of the local
governments. Hence, the SHGs came to be accepted as the ideal institutional
structure for the implementation of development projects constituted under WCP
(Isaac el al 2002). Also, the plan guidelines permitted financial assistance to selfhelp groups in the form of revolving funds.
The Kudumbashree initiative of the Government of Kerala is characterised as a
'women-<>riented' poverty eradication programme with women's empowerment as
one of its major goals - one that is actively projected in the media. Kudumbashree
places women from BPL households at the heart of the poverty eradication efforts of
the State. It was begun in 1998 by the Government of Kerala with the aim of
eliminating poverty within ten years, by 2008. The present structure, however, began
to evolve earlier, in 1991, when the Community-Based Nutrition Programme
(CBNP) was initiated by the Government of Kerala with active help from UNICEF
to improve the nutritional status of women and children. In this initiative, threetiered structures made up of neighbourhood groups (NHGs) that were federated into
area development societies (ADS) at the ward level, which were in tum federated
into community development societies (CDS) at the municipal level, were formed,
composed exclusively of women from families identified as poor through a non-

35

l
income based indexl. The success of the CDS model in urban Alappuzha and in
rural Malappuram led the Government of Kerala to scale up the strategy to the whole
of the State in 1998 under the name Kudumbashree, with the State Poverty
Eradication Mission taking the responsibility of implementation through the
Department of Local Self-Government. The Kudumbashree has since followed a
mUlti-pronged strategy which includes convergence of various government
programmes and resources at the community-based organisation, efforts to involve
the CDS structure in local level anti-poverty planning, and development of women's
micro-enterprises and thrift and credit societies. While micro-credit is one of the
many strategies initiated by Kudumbashree, it has been gaining importance and
visibility within the programme,.
[t is discernable from the above discussion that recent efforts made by the State to
increase economic visibility of women has certainly changed the economic lives of
poor women. Earlier, paid employment in the organised and unorganised sectors was
the main source of employment for women. [n addition, self-employment was a
probable way out, but was constrained by lack of sufficient capital in the hands of
poor women. Formal lending institutions also become inaccessible for them due to
lack of sufficient collateral. Hence, self-help groups of women became a solution to
the above problem and it has 'become an accepted organisational mechanism to
coordinate various anti-poverty programmes including the WCP. Another feature of
A nine-point non-income criteria developed in the Alleppey CDS has been adopted by
Kudumbashree, since Alleppey has turned out to be a successful experiment. For a detailed
description, see Isaac e/ 01 200 I.
2 The CDS model in Alleppey was introduced in 1993 with the collaborative efforts of Community
Based Nutrition Programme (CBNP) of UNICEF and the Urban Poverty Alleviation Project (UPAP)
ofGovemment offndia.
1

36

micro-credit is that it bundles a range of financial activities, viz. thrift savings, thrift
loans, linkage with commercial banks etc. Moreover, it does not stand as a financial
instrument alone. As members of the kudumbashree, women get involved in the
local planning process and often become beneficiaries of welfare measures. As a
result, participation in kudumbashree at varying levels has emerged as an inevitable
element of the economic as well as the social life of poor women.
3.3. Methodological issues in women's work
The study context of understanding women's work has been described in the above
section. It is evident that women's work is slowing getting reoriented towards microfinance activities.

However, these activities are of varied nature and there are

specific aspects pertaining to wQmen's work. The following is a description of such

specificities of women's work and the methodology adopted in the present study to
understand them.
3.3.1. Undercounting of women's work
Recognition of the fact that much of women's work in the developing world is
overlooked, undercounted, and undervalued is not new. It is well established that in
primarily capturing paid employment, statistics on women's work ignore unpaid
domestic labour, subsistence p~uction, family farm work, and informal incomegenerating activities. For the last few decades, researchers from different fields have
described the inadequacies of official labour force statistics, questioned traditional
notions of what is and what is not work, and designed and implemented better

37

methods of data collection (Boserup 1970; Beneria 1982; Dixon 1982; Tinker 1990,
Anker and Anker 1995; Donahoe D A 1999).
Women work both for the labour market and for their households. However, the
major criticism is that women

VIlIt0 work in activities which are commonly perfonned

by men are much more likely to be counted as working. Dixon (1982) has referred to
women as "invisible workers" because of the firilure to enumerate domestic labour in
the census and in the national accounts mainly because it is of use value rather than
exchange value. Desai and Jain (1994) in their study on rural South India noted that
with the supposed contlict between children's need for care and mothers need for
income, the role of the domestic labour of women tends to be ignored. In addition to
domestic activities like cooking, cleaning etc., women spend considerable time for
other activities like the making 9f cow-dung cakes, collection of firewood etc., which
are rarely considered economic activity. Although they are highly productive, they are
called "marginal economic activity" or "expenditure-saving activity" (Desai and Jain
1994).

The necessity for more comprehensive measures of women's work to accurately

describe their productive work is clear. A precise portrait of women's work is
necessary to reveal women's hidden economic activities, recognize the extent of
their contribution to family well-being, fully understand the nuances of their social
position, and make infonned oevelopment interventions (Tinker 1990). While an
extensive literature documents the need for better measures of women's work,
relauvely few attempts have been made to construct suitable work typologies.

38

3.3.2. Multiplicity and Intensity of work

A significant aspect of economic life takes place in the area of production, largely
ignored by standard macroeconomic analysis, namely the household production of
goods and services. Rural households produce a wide array of goods and services
for their own use and consumptton, including growth of subsistence crops, gathering
water and fuel, food preparation and house cleaning, care of children and elderly etc.
These non-marketed goods and services are vital for social reproduction and human
development. Moreover, there is dynamic interaction between household production
and production for the market economy, as household members, especially women,
allocate their time between the two sets of economic activities. Also for them,
engagement in work - whether production for their own consumption, for the
community at large, or for the market - constitutes an essential element of life.
Numerous studies tend to exp1J!in the decisions relating to allocation of women's
time among these various activities using a host of determinants, such as market
wages, life cycle of the household, household wealth status, women's schooling and
age etc. (Floro, 1995) There are, however, two important areas in the literature which
have received scant attention is the qualitative dimension of time use, i.e. the
intensification of work time and the impact of such work on women's well-being and
that of their children (Floro, 1995).
Women undertake multiple tasks that fall within the arena of production for the
market as well as for the household consumption of goods and services, and often
these tasks overlap. Given the difficulty of measuring this qualitative dimension of
time use, the extent to which women perform certain overlappi ng activities serves as

39

an indicator of the intensification of work time or work intensity. It has been argued
that simultaneous performance over prolonged periods of market or domestic
activity, or more than one type of activity, increases the intensity of work performed
by the individual and that the overlap of leisure and work activities adds a work
dimension to what is generally considered leisure (non-work) time.
There is growing evidence that the performance of overlapping activities over
prolonged periods, particularly by women, is not an isolated phenomenon. Robinson
(1985) and Hill (I 985} employ time-use surveys of US men and women in the mid1960s, 1970s and early 1980s to examine the presence of overlapping activities. By
comparing two methods of recording time use, namely, time diary keeping and
survey questionnaires, these studies argue that the discrepancy in time-use results
can be largely explained by overlapping of activities. Their finding also suggests
that the incidence of performing more than one task at a time is gender-specific, that
is, it is more prevalent among women than men, particularly with respect to
performing domestic chores. Several studies on home working and the practice of
'putting out' in certain informal sector activities, indicate the strong likelihood of
women combining paid work and domestic activities such as cleaning, cooking and
childcare (Beneria 1992; Lozano 1989; Christensen 1989). Such jobs are essentially
industrial tasks that are carried out in a residential context.

Although income

deriving from home working is very low, in most cases below the minimum wage,
flexibility is the most important reason for women opting for both home-based
wor~ing and informal sector ac~ivities. Certain types of informal sector activities

also allow women to reconcile their reproductive and productive roles. In a Delhi

40

study on slum dwellers, Karlekar (1982) points out that 30"10 of poor women who
work as street sweepers end up taking their children with them to the workplace.
The study result suggests that the incidence of combining childcare with market
activities is likely to be higher for women with fewer childcare substitutes available.
Hence, the above studies suggest that women's work needs to be understood in its
context and a suitable methodology needs to be devised in order to comprehend it
meaningfully. Time-use survey ;md focus group discussions are employed to develop
an appropriate typology of work in the present study and it has been described in the
following sections.
3.3.3. Time use survey method to measure women's work
The most preferred method to understand both market production and non-marketed
household production of goods and services is time-use surveys in spite of its
methodological weaknesses.

One criticism often made against is that time use

surveys tend to measure only drrect market use of time, which leaves a great bulk of
human activity unmeasured, namely, the time spent on non marketed production and
leisure activity. Moreover, the reported hours of work appear to be substantially
over-stated relative to actual hours in the work place (Juster and Stafford 1985).
There have been several time-use studies focusing on women's work undertaken in
developing countries. Alternative methodologies are used by such studies to measure
the labour input of various economic activities (Juster 1985; UN I 990b). These
include direct participant observation, the so-called stylized survey method, the
random-hour method and time-diary method (Floro, 1995). Though the direct

participant observation is the ideal method, it might have an effect on the work
behaviour of the respondent. Observation of a set of activities may tend to alter the
nature of the activities themselves, since the observer may be perceived as an
'intruder'. Moreover, it involves greater cost in collecting information. The random
hour method employs time-use reports obtained from respondents at randomly
selected moments during the same day. While this is not as costly as other methods,
there is the obvious problem of activities, which are highly irregular. Another form
of time-use survey involves the..measurement of time use, based on a "stylized" list
of activities. A series of questions inquiring first about the frequency of and then the
typical duration of the activity is used. Such a method allows for a rapid appraisal.
However, it poses the basic problem of perception in the respondents' ability to
provide accurate information. There is growing consensus among researchers that
the preferred method of data collection on time use involves a chronological record
of various activities over the day through the use of a diary instrument. Time-use
diaries are preferred over the other methods for they tend to be more comprehensive,
enable respondents to report a<;tivities in their own terms and have some form of
built-in check that increases the reliability of such data (Juster 1985). Despite these
clear advantages, there may be a number of deficiencies that can lead to serious bias
in the results. Unless one is able to sample days that encompass all important
variations in time allocations, such as seasons, weekends, holidays etc., there are still
problems of inaccuracy.

3.4. The nature and specificities of women's work in the study setting
In the light of the above discussion on methodological issues involved in measuring
women's market as well as non-market work, the present study employs a multiple
method that includes both time-use survey and focus-group discussion. Since the
sample for the study has been selected from families below the poverty line, all of
them are members of the three-tier self-help group structure federated at the local
village panchayat level. Hence, the focus group discussions were carried out in the
weekly meetings of self-help groups of women members. Questions on their type of
work, time management, selectivity of work, leisure etc. have been discussed. Also,
they were asked to list all activities in a typical day, as well as the allocated time for
the same. Both the focus group discussions and the time use survey have provided an
apparent picture of the nature of women's work in the study area and a description of
the same is given below.

.
of work
Table 3 2Pattern
Type of work (main Income eammg
activity)
Household domestic work alone
Helpers in construction work
Domestic servants
Fish and Vegetable Vendors
Home based workers
Kudumbashree Micro Enterprises
General wage workers
Self employed
Petty shop
Others
Total

Percentage
distribution
43
6.1
4.1

5
2.5
18
5
7.3
2
7
100

43

The above table shows that 57 % of women are engaged in some fonn of incomegenerating activity. Though initially fewer number of women have been reported as
working, a further probe revealed that many of them are engaged in a variety of
income-generating or expenditure saving activities such as poultry keeping, dairy,
maintaining a kitchen garden, tailoring for the family etc. One reason might be that
the peoples' planning campaign, which was experimented within Kerala during the
period 1997 - 200 I, has promoted these activities through giving them productive

assets and other incentives. In addition to this, activities such as sending children to
school, collecting water from the public standpipe etc. are some of their time
consuming activities.
A multiplicity of work has been found in the study field. In general, women do
domestic work as well as provide market labour and indulge in self-help activities in
varying degrees. The self-help activities are at two levels. Some of the respondents
are only members of the group and are involved in the activities of the group such as
weekly meetings, collecting thrift, providing thrift loans and other development
activities. Whereas some other members set-up micro enterprises as a group-based
activity and carry out income-generating activities.
Another aspect is that many of the respondents who are reported to be engaged in
economically productive work are involved in more than one activity. For example,
one woman has reported that she is a helper in the construction sector. Whenever
there is no work in the construction field, she works with her mother who is working
in a brick kiln. This kind of involvement in different productive works as well as

44

productive and reproductive work is found. Hence the multiplicity in the work is one
of the specificities of women's working lives in the study setting too.

The intensity of work is yet another dimension of women's work. For those who are
engaged in market work, if there is anyone at home to share their domestic and other
related work, the intensity of work seems to be less.

However, a considerable

number of them manage different tasks single-handedly and its impact on their own
physical and mental health is an important aspect, which calls for scholarly concern
and policy attention.

When compared to the labour markets of the western countries, in our country the
process of entry into and withdrawal from the labour force is not very flexible.
However, it was found that many of those who were engaged in market production
have completely withdrawn from the labour market after marriage and a sizable
proportion of them have re-entered after childbirth. Hence, there is dynamic
interaction between women's involvement in the market production and the
production of non-marketed goods and services for social reproduction.

Generally, the categorization of mothers as working and non-working is made while

analyzing the impact of their work on child health and care under the assumption that
non-working women can spend more time with their children. However, it is found
that this assumption of non-working women as a homogeneous category in terms of
spending more time on childcare is wrong. There is a flaw in this assumption and it
largely depends on factors such as the strul:ture of the household (nuclear or
, extended), pattern of sharing of domestic work with other members of the family and
\

45

availability of other modem time saving facilities at home. Time-use survey has
found that the amount of time women spent on childcare and other various types of
work they perform, including domestic work varies based on the number of activities
they perform. Women in generai do domestic work, take part in self-help activities,
run micro enterprises under self-help groups and engage in market labour in varying
degrees. It has been found that the more the number of activities they perform, the
less the amount of time they spend on childcare.
Therefore the following work typology has been adopted based on actual time use.
I. Performing domestic work and self-help group activities (DW+SHG)

2. Performing domestic work, self-help activities and engaging in micro-enterprise

production (DW+SHG+ME)
3. Performing domestic work, market labour and self-help group activities
(DW+ML+SHG)
4. Performing domestic work, market labour, self-help group activities and engaged
in micro enterprises activities (DW+ML+SHG+ME)

(Ow - domestic work; ML - market labour; SHG - self-help groups activities; ME-

micro enterprises)

46

Table 3.3: Time spent on work and child care bv typolo2Y of work
Time* spent Time* spent
Frequency Percent
Typology of work
on different
on childcare
works (hrs)
(hrs)
181
43
6.46
DW+SHG
6.37
61
14.5
8.50
5.37
DW+SHG+ME
DW+ML+SHG
107
25.4
12.05
4.29
72
17.1
14.20
DW+ML+SHG+ME
3.8
421
100
Total

* Average time in hours in a day

The above table shows that 57% of women are involved in any of the incomegenerating activities. The total time spent on work increases along with increase in
the number of tasks. A subsequent reduction is found in the time spent on childcare.
Compared to women who are combining all selected categories of work, i.e.
performing domestic work, market labour, self-help group activities and running
micro enterprises, women doing only domestic work and self-help group activities
spent almost double the time on childcare. However, this does not provide a detailed
picture of the distribution of maternal time across various caring activities. The
activities in the childcare regime and the amount of caring work being taken up by
all members of the family is discussed in detail in chapter five.

Though the amount of time that mothers spend with children is reduced when they
engage in more than one income-generating activities, whether such maternal work
adversely affect child health is ,the research question that has been examined, using

bot secondary and primary data sources in the following two chapters.
\

47

3.5. Summary
In the present chapter the context in which the working lives of women are situated
is examined. It is found that there is a considerable degree of economic restructuring
of work around micro-credit activities taking place among poor population in Kerala.
This state led initiative also made effort to link up women-oriented self-help groups
to local level PRls. As a result, the notion of women's work is getting redefined
mainly around the micro enterprise production. A detailed discussion on conceptual
issues involved in understanding women's work in developing countries such as
undercounting, undervaluing, intensity and mUltiplicity of work. The same has been
analysed in the study setting using both primary quantitative and qualitative data. To
overcome these shortcomings, an appropriate typology of work has been devised
using time use survey. It is found that as the number of activities on the work front
"

increases, the amount of time that mothers spend on childcare decreases. To

understand this further, a detailed study of childcare regime and the activities that
women perform etc. is necessary and this is done in chapter five.

48

Chapter 4

An analysis of the determinants of child health using the

National Family Health Survey Data
4.1. Introduction
One of the objectives of the pr\lsent study is to understand the causal relationship
between maternal employment status and child health outcomes. Some studies in the
Indian context (Basu and Basu 1991; Kishore 1992) have suggested that women's
employment outside the home may have at least one disadvantage: the survival of
young children appears to be negatively affected if mothers women work outside the
home. However, studies in general have shown mixed evidence on the direction of
causal relationship between maternal employment and child survival. Past research
investigating the association between women's employment and child mortality
using 1981 Census data has follnd specifically, at the district level, higher rates of
female labour force participation have found to be associated with higher rates of
child mortality net of several cultural and economic factors (Kishore 1992; 1993).
Whereas, another study using the same data set found no such significant association
between female labour force participation and child survival (Murthi, Guio and
Ore'ze 1995). There have been follow-up studies by Mamtha Murthy and Jean
Ore'ze dealing with later evidence, particularly the 1991 census, which broadly
confirm the findings based on the 1981 census2 Hence, it still remains a gray area of
~h that requires further empirical investigation as well as reflection .

..

,See Jean Dre'ze and Mamtha Murthy, "Female Literacy and Fertility: Recent Census Evidence from
.

India", mimeographed, Centre for History and Economics, King's College, Cambridge, UK, 1999

49

In the present chapter, an analysis of the secondary data i.e. National Family Health

Survey (NFHS) on Kerala is made in order to understand the causal relationship

between women's work and child health. The NFHS data provide information on
certain child health indicators. Nevertheless, it does not provide any information on
various aspects of childcare and investment of time on childcare by mothers as well
as other familial members. Hence, the multivariate analysis attempted here has
"

primarily used morbidity incidence among children as an indicator of child health.

A cross-tabulation of selected child health indicators by background characteristics is
also provided before attempting multivariate analysis.
, 4.2. National Family Health Survey Data
The National Family Health Survey (NFHS) is an important household survey
launched by the Ministry of Health and Family Welfare (MOHFW).

It was

undertaken with the principal objective

of providing state-level and national level
,
,

estimates of fertility, infant and child mortality, the practice of family planning,
maternal and child health care and the utilisation of services provided for mothers
and children.

The NFHS covered 24 states and the national capital territory of Delhi, which
comprise 99 percent of the total population of India. The NFHS I in Kerala, which
was conducted between October 1992 and February 1993 gathering information
from all the fourteen districts on a representative sample of 4332 ever-married
women in the age group of \3 -'49 from 4387 households. The survey also collected

50

extensive health-related infonnation on 1605 children born to these women during

the four years preceding the survey.
The second round of NFHS was carried out in 1998-99. It covered a representative
sample of about 90,000 eligible women in the age group of 15-49 from 26 states that
comprise more than 99 percent of total population of India. In Kerala, NFHS II
collected infonnation from 2834 households between 22 March and 20 July 1999
and interviewed 2884 eligible' women in these households.

Also, the survey

collected infonnation on 707 children born to eligible women in three years

preceding the survey.
The following is cross-tabulated data from the NFHS II round of certain aspects of
child health viz. prevalence of morbidity episodes, nutritional status, and anemia
among children, by various background characteristics including maternal work
status.

It broadly provides a picture of the relative significance of certain

background characteristics that might influence child health outcomes.

51

Table 4.1: Prevalence of acute respiratory infection, fever and diarrhea by

back2round characteristics
Cough
with fast
Total
Fever
Diarrhea
Background characteristics
breathing
(ARI)
Religion
303
10.3
42.6
23.1
Hindu
280
41.6
12.8
Muslim
22.4
Christian
103
38.2
12.2
23.3
Caste
54
55.7
16.4
Scheduled caste
31.8
Other backward caste
294
36.4
12.5
23.4
Others)
331
20.3
43.7
10.3
Sex of the Child
44.0
Male
25.4
14.9
352
Female
335
20.1
38.9
8.1
Education
Illiterate
(24.6)'
(24.9)
(18.2)
29
Literate abv. middle school
24.3
39.8
11.4
148
Middle school complete
26.8
48.5
13.6
142
High school and above
20.6
40.8
10.4
368
Residence
Urban
130
23.0
36.3
12.7
Rural
22.8
42.7
11.3
556
Standard of living index
Low
11.1
21.8
41.4
98
Medium
24.7
43.0
12.3
388
High
19.7
38.8
10.5
201
Mothers work status
Employed by some one else
25.9
46.9
15.3
62
Not worked in the past 12
24.4
40.3
11.1
612
months
Source of drinking water
Piped water
23.4
44.3
15.3
121
Well water
23.3
41.2
11.0
538
(Source: NFHS II Report on Kerala, 200 1)

, Not belonging to a scheduled caste, a scheduled vibe or an other backward class

( ) Based on 25-49 unweighted cases

52

The above table shows that the prevalence of all morbidity episodes are high among
the scheduled caste community in comparison to others. Also compared to female
children, male children are more affected by the three illnesses. This might be due to
both the biological disadvantages of boy children in the younger age groups and the
high degree of exposure to the environment. It also shows that the percentage of
morbidity episodes among children decreases only above the middle school level of
maternal education. It is evident that the experience of illnesses is high among
children of mothers who work compared to those who have not worked in the past 12
months. Whether such maternal employment significantly determines the child
morbidity has been analysed in the following section using multivariate logistic
regression.

Children from households with piped water supply have higher

morbidity prevalence than those with well water, which is more prominent in the
case of diarrhea, a water-borne disease. Chances of the water getting contaminated
are more in the case of piped water as compared to well water. In the case of
morbidity incidence, children belonging to households having a low or medium
standard of living are similar whereas it is less in the case of children from
households with a high standard" of living.

53

Table 4.2: Nutritional status ol" children by bacliground characteristics

Weight for
Height for age
Background Characteristics Weight for age
height
Percent>-2S0 5 Percent>-2S0
Percent>-2S0
Religion
20.7
13.0
29.6
Hindu
10.5
25.8
Muslim
27.4
Christian
15.4
13.4
6.7
Caste
(12.3)
(43.0)'
Scheduled Caste
(38.21
Other backward caste
30.6
23.4
12.5
Others6
20.1
17.7
9.2
Education
Literate abv. middle school
30.5
25.0
18.3
Middle school complete
31.2
24.6
8.6
High school and above
23.6
19.3
9.5
.
Residence
Urban
22.4
18.5
10.9
Rural
28.0
22.7
11.2
Standard of Living Index
Low
35.8
27.8
16.2
Medium
28.8
24.4
11.2
High
19.2
14.4
8.8
Mothers work status
Employed by some one
(24.0)"
(21. 7)
(4.0)
else
27.2
22.5
11.6
Not worked in the past 12
months
Total
26.9
21.9
11.1
.

(Source: NFHS II Report on Kerala, 200 I )

Weight for age, height for age, weight for height are the three anthropometrical
measures selected to measure the nutritional status among children. It shows that the
nutritional status among children in terms of selected indicators is better in urban

compared to rural areas. It indicates that children of mothers with education of

'The index is expressed in standard deviation units (SD) from the median of the International
Reference Population.
'Not belonging to a scheduled caste, a scheduled ~be or an other backward class
, ( ) Based on 2549 unweighted cases
.
Based on 2549 unweighted cases

54

the level of high school and above are better off in terms of selected nutritional
indicators.

In comparison to tlhildren belonging to Hindu and Muslim families,

children from Christian households have better nutritional status. It also shows that
children belonging to the SC community are nutritionally weak compared to children
from the other two communities and children from the forward community are
nutritionally better positioned in comparison to children from the OBC group.
Children of working mothers are slightly better off in nutritional status compared to
the children of mothers who do not work. There is a direct positive correlation
between standard of living of the household and the nutritional status of the children.

55

I ren
Table 4.3: Anaemia among c b'ld
Background characteristics

Religion
Hindu
Muslim
Christian
Caste
Scheduled Caste
Other backward caste
Others 1O
Sex of tbe Cbild
Male
Female
Education
Literate abv. middle school
Middle school complete
High school and above
Residence
Urban
Rural
Standard of Living Index
Low
Medium
High

Percentage of chi ldren with

9
anv tvoe of anaemia

44.4
41.8
48.9
(36.1)"
38.6
49.2
42.3
45.7

(32.9)
46.0
45.4
46.8
43.2
42.1
44.7
43.3

(Source: NFHS II Report on Kerala, 200 I )The table 4.3 shows that overall 44
percent of children are anaemic. The prevalence of anaemia is high among female
children and children belonging to scheduled caste or other backward communities.
The prevalence of anaemia is lower among children whose motbers are employed by
someone else compared to the children whose mothers have not worked in the past
12 months. Maternal education does not show any correlation with anaemia among
their children.

9 Any type of anaemia includes mildly anemic (10.0-10.9 gldl). moderately anaemic (7.0-9.9 gldl) and
severely anaemic (less than 7.0 gldl)
.
10 Not belonging to a scheduled caste, a scheduleo tribe or an other backward class
11 Based on 25-49 unweighted cases

56

4.2.1. Analysis of the determinants child health using NFHS I data

It has been seen in the above discussion on child health indicators by background
characteristics that the incidenIJe of child morbidity is found to be more among
children whose mothers are employed. Hence, multivariate analysis has been carried
out to find the determinants of child morbidity and the impact of maternal work
status. For this purpose, the incident of diarrhea among children during the reference

period of the two previous weeks of the survey is selected as the dependent variable.
Since the dependent variable is dichotomous in nature, logistic regression has been
adopted to quantify the net effect.

In a logistic regression, the coefficient of the ith variable measures the increase or
decrease in the log of the odds of the event of interest occurring for every one unit
increase in the ith variable. Thus, if p is the probability of morbidity incidence, the
event of interest in this case, then 0 i gives the change in the value of Log (P/I- p)
when ith explanatory variable increases by one unit. For ease of interpretation, the
odds ratios are given in the table 4.5, rather than coefficients.

The educational status of the mothers, type of sanitation facility available, type of
drinking water facility, source of lighting, type of household, area of residence and
sex of the child have been seh,cted as the explanatory variables in addition to the
work status of mother.

57

. bl es seI ec ted
f be explana t ory vana
Table 4.4: Detal'1sot
Value
Explanatory variables
Religion
0
Hindu
Muslim
I
Christian
2
Caste
Scheduled Caste
0
Other backward caste
I
Others l2
2
Sex of tbe Cbild
Male
0
Female
I
Education
Illiterate
0
Literate abv. middle school
I
Middle school complete
2
High school and above
3
Residence
Rural
0
Urban
I
Motbers work status
Not working
0
Working
I
Source of drinking water
Others
0
Public pipe/well
1
Own pipe/well
2
Type of housebold
Kutcha
0
Semi-pucca
1
Pucca
2
Source of lighting
Others
0
Electricity
1
Sanitary facility
No toilet facility
0

Public toilet
I
Own toilet
2

Frequency

549
189
162
72
423
405
447
453
107
260
257
276
685
215
711
189
19
294
587
220
448

232
173
727
88
165
647

" Not belonging to scheduled caste, scheduled tribe or \lOy other backward class

58

Table 4.5: Incidence of diarrhea among children on selected explanatory

variables Odds ratios
Dependent variable:
Explanatory variables
Incidence of diarrhea
Religion
Hindu rc
Muslim
1.08
Christian
1.01
Caste
Others!3 rc
Scheduled Caste
1.22'"
Other backward caste
\.12"
Sex of the Child
Male rc
Female
0.96
Education
Illiterate rc
Literate abv. middle school
0.98
Middle school complete
0.89
High school and above
0.77'
Residence
Urban rc
Rural
1.03
Mothers work status
Not working rc
Working
1.05
Source of drinking water
Own pipe/well rc
Public pipe/well
1.24"
Others
1.36'"
Type of household
Kutcha rc
Semi-pucca
0.78"
Pucca
0.70'"
Source of lighting
Electricity rc
Others
\.18'
Sanitary facility
Own toilet rc
Public toilet
1.26***
No toilet facility
1.83*"
-2Ioglikelihood: 549.43

Constant : 1.78

***p<.OI; **p<.05; *p<.IO

" Not belonging to a scheduled caste, a scheduled tribe or an other backward class

S9

(Note : The odds ratios have been presented above and in the logistic regression,
odds ratio is interpreted as the proportionate change in the odds of the event
occurring for a unit change in the value of predictor variable. The odds ratio of the
reference category is one by definition.)
The results shown in the above' table 4.5 reveal that maternal education and work
status are not significant in detennining selected child health outcome. Whereas
variables indicating socio-economic status of the household and infrastructure
facilities turn out to be significant in detennining the morbidity. Also the caste status
turned out to be significant in detennining child morbidity. Same analysis has been
carried for the other two types of morbidity i.e. acute respiratory infection (ARI) and
fever. The results yielded a similar pattern in the case of ARI whereas in the case of
fever, except area of residence, all the explanatory variables were found to be
insignificant and hence are nof' presented here. Following is the similar analysis
carried out with NFHS II data and a discussion of the results of both the rounds of
NFHS is given below.
4.2.2. Analysis of the determinants of child health using NFHS II data
The analysis carried out with NFHS I has been repeated using NFHS II data. For
this analysis, an indicator of child health i.e. the incidence of diarrhea among
children during the reference period of the previous two weeks of the survey is
Jelected as the dependent

varia~le.

Since the dependent variable is dichotomous in

. nature, logistic regression has been adopted to quantify the net effect.

The

explanatory variables selected are: (1) the educational status of the mother, (2) work

60

status of the mother, (3) caste, (4) type of the sanitation facility available, (5) type of
drinking water facility, (6) source of lighting, (7) type of household, (8) area of
residence, (9) sex of the child. (10) female autonomy.
An additional explanatory variable considered in the regression analysis is female
autonomy. Female autonomy has been measured in NFHS (II) in terms of household
'.

decision-making power, mobility status and access to the resources, The decisionmaking variables are the following: (I) who decides what to cook, (2) who decides
on obtaining health care, (3) who decides to purchase jewelry, (4) who decides about
the respondent staying with the family. The options given are: (I) respondent, (2)
husband, (3) others in the household, (4) jointly with the husband, (5) jointly with
others in the household. The mobility indicators are measured in terms of: (I)
permission needed to go to market, (2) permission needed to visit relatives or friends,
and the given option is yes or no, The access to resource is measured by asking the
woman whether money is allowed to be set aside and the given option is yes or no.
A composite index of autonomy is constructed by assigning values ranging 0 to 1 for
all the indicators and taking a sum total value.

For decision-making variables,

'respondent' is given a value of I and 'husband' and 'others in the family' are given
a value of 0 each. 'jointly with husband' and 'jointly with others in the family' are
given 0.5 each. For the rest of the variables, a value of 1 is given for 'yes' and 0 for

'1

0"

Hence, the autonomy index ranges from 0 to 7.

The following are the details of the variables selected.

61

. bles se ected
Table 4.6: Details 0 f t he exp anatorv varIa
Value
Explanatory variables
Religion
0
Hindu
I
Muslim
2
Christian
Caste
0
Scheduled Caste
I
Other backward caste
l4
2
Others
Sex of the Child
0
Male
I
Female
Education
0
Illiterate
Literate abv. middle school
I
Middle school complete
2
High school and above
3
Residence
0
Rural
I
Urban
Mothers work status
Not working
0
Working
I
Source of drinking water
Others
0
Public pipe/well
I
Own pipe/well
2
Type of household
Kutcha
1
Semi-pucca
2
Pucca
3
Source of lighting
Others
0
Electricity
1
Sanitary facility
No toilet facility
0
Public toilet
I
Own toilet
2
Female Autonomy
>2
0
1
~4
2

I.

Not belonging to a scheduled caste, a scheduled tribe

FreQuency
437
125
131
49
326
319
350
344
77
205
198
214
485
209
570
124
127
464
103
164
382
148
498.
196
111
463
120
298
357
68

or an other backward class

62

Table 4.7: Incidence of diarrhea among children on selected explanatory

variables - Odds ratios
Explanatory variables
Dependent
variable:lncidence of
diarrhea
Religion
Hindu rc
Muslim
1.12
Christian
0.98
Caste
Others J5 rc
Scheduled Caste
1.32***
Other backward caste
1.17**
Sex of the Child
Male rc
Female
0.95
Education
Illiterate rc
Literate abv. middle school
0.99
Middle school complete
0.86
High school and above
0.79
Residence
Urban rc
Rural
1.03
Mothers work status
Not working rc
Working
1.02
Source of drinking water
Own pipe/well rc
Public pipe/well
1.53**
Others
1.27**
Type of household
Kutcha rc
Semi-pucca
0.75**
Pucca
0.63**
Source of lighting
Electricity rc
Others
1.31 **
Sanitary facility
Own toilet rc
Public toilet
1.24***
No toilet facility
1.56***
Female Autonomy
>2 rc
2 4
1.12**
<4
1.22***
"N

'

'

01 belonging to a scheduled caste, a scheduled tribe or an other backward class

63

-2 log likelihood: 432.87

Constant: 1.34

"p<O.OI; "p<O.05; p<O.1O

(Note : The odds ratios have been presented above and in the logistic regression;
odds ratio is interpreted as the proportionate change in the odds of the event
occurring for a unit change in the value of predictor variable. The odds ratio of the
reference category is one by definition).
Similar analysis has been carried out with the two other indicators of child
morbidity, viz. incidences of acute respiratory infection and fever as dependent
variable. The results yielded a similar pattern in the case of ARI. Hence the results
are not presented here. In the case of incidence of fever, most of the selected
explanatory variables were found to be insignificant.

4.3. Discussion
The results of the logistic regression analysis of both the rounds of NFHS data show
that compared to individual level variables such as maternal education and work
status, variables representing household socio-economic characteristics such as
sanitation, drinking water etc. turned out to be significant in explaining morbidity
incidence among children. Also children belong to backward communities are more
likely to have morbidity incideni:e when compared to children offorward castes.
In the analysis, maternal education was found to be very weak in explaining
morbidity incidence among children.

In fact, in the demographic research, the

positive impact of maternal education on child health is well researched and

tocumented. It has been argued that educated mothers are better able to access
medical services and follow advice; they have greater power within the household to
ensure that sick children receive care and

~ey

engage in childcare practices thaf

enhance child health. Hence, there is increased importance of girls' schooling in the

64

'.

realm of policy formulations. In fact, this better educational status of women always
acts as a proxy of the socio-economic context in which they live, especially in the
setting of developing countries. Advocating girls' schooling for better child health
has certain inherent problems. On the one hand, it considers educating girls - future
mothers - as an important pre-requisite to enhance child health and thereby shifts the
burden of dealing with childcare (including dealing with diseases) from the society
as a whole to women in poor households. The problem here is that the investments
in girls' education is not articulated on the grounds of equality and as a matter of
right. This may be viewed in the backdrop of globalisation and the declining trend in
investments in public health. It diverts attention from the need to invest in public
healthcare facilities such as sanitation and water supply. Thereby, it caters to the
ideology of nea-liberal institutions by focusing on the role of mothers in enhancing
child health and shifting the boundaries between public and private responsibilities
for care of children.
Recently, there has been a policy shift in women's employment generation towards
self-help group activities. Self-~elp group activities are more of self-employed work
and there is no social security attached to the work. In addition, there are family
health insurance packages providing for the poor women of SHG groups. In the light
of the above discussion on shifting the boundaries of social responsibilities, the
impact of these strategies has to be critically examined.
Quite contrary to our expectation, it is found in the analysis ofNFHS II data that the
higher the levels of female autonomy the more is the likelihood of child morbidity.
It basically questions the conceptual framework which suggests that the autonomy
65

women achieve through their ea:tployment outside home would help to improve the
health status of their children. Therefore, it is important to have an understanding of
the factors that determine female autonomy as measured in NFHS II and its different
dimensions. Such an analysis has been carried out in chapter six.
4.4. Summary
An analysis of factors determining selected indicators of child health (such as

incidence of diarrhea among children) is carried out in the present chapter using
NFHS I&II. The results indicate that the often-highlighted individual variables in
the demographic literature such as female education and work status in the
demographic literature do not have significant influence on child morbidity
outcomes. Living conditions reflected by water supply and sanitation facilities tum
out to be significant factors. The implications of this result in a context that exhibits
consistent reductions in public healthcare investments and policy advocacy for
women's schooling and employment are discussed.
The following chapter also analyses the relationship between maternal employment
and child health outcomes using primary data collected from the study field. It also
conceptually explores various dimensions of childcare and the impact of maternal
work on time spent on childcare.

66

Chapter 5
The impact of women's work on child health and care-A
quantitative and qualitative exploration of primary data
5.1. Introduction
The impact of women's work on child well-being is assessed in the present chapter,
using primary data collected from the study field. The method of analysis includes
both qualitative and quantitative inquiries. Using primary a survey instrument,
information has been collected on various aspects of childcare. It covered 590
children out of which 42

percen~

belong to the 0-3 age group and 58 percent belong

to the 4-7 age group. The entire activities of the child on the day previous to that of
the day of data collection have been traced and mapped on a time scale of 24 hours
and the persons engaged with the child during these activities have been mapped
along with the duration of such time. This information provides an idea about the
regime of childcare. However, since there are many caring chores that need to be
done without the presence of the child such as washing clothes its clothes, cooking
special food etc., it does not provide complete data on all childcare activities. Hence,
separate questions have been a;;ked of the mothers about the amount of time they
spent on childcare. Childcare involves an array of activities with varying time inputs.
In order to understand the regime of childcare properly and to make a list of caring
a~tivities, it has been included in the focus group discussions about various caring

activities. The qualitative information collected in the focus group discussion has
also been used to validate the questions asked in the primary survey.
\

67

The following activities have been listed by participants in the focus group
discussion as the components of childcare that require time inputs.

I. Feeding

2. Cooking meals

3. Bathing
4. Washing clothes
5. Changing their clothes

6. Playing
7. Carrying/holding

8. Attending when children cry and comforting them

9. Special cleaning of the premises

10. Taking children to health care centers
II. Providing special care and attention when they are sick

12. Extra care on personal hygiene especially when children are young
\3. Spending time on talking to children to enhance their verbal abilities

14. Assisting children in studies

15. Making children sleep
16. Overall supervision of all caring activities done by others
S.2. Methodological issues involved in the measurement of childcare time

lIhe following

is a description of methodological issues involved in measuring

childcare.

68

As discussed in the case of women's work, in childcare too intensity and multiplicity
of activities can be observed. Also, there is active and passive caring time.
5.2.1. Concept of active and passive care and time demands
Conceptualization of caring time is an important aspect in the methodology of
quantification of childcare time. Most of the time-use surveys ask adults about their
participation in activities with children, overlooking the demand for supervisory time
or "passive care". The passive forms of care broadly include monitoring children
outside playing or sleeping, preserving a safe environment, being an adult presence
for children to turn to in need, supervising games etc. (Folbre et aI, 2005). Survey
designs in general focus on parents or family members living in the same household,
rather than the total amount of ~npaid care provided by family members, including
those who are not living in the same household.
5.2.2. Intensity and multiplicity of activities in child caring
The multiplicity and intensity of activities is a ml\ior characteristic of childcare.
Women do cooking and other household work, carrying children around and feeding
them, do home-based income earning work etc. simultaneously. Often, women
combine child-caring activities with their leisure. Hence, it often happens that they
do the same quantity of work in' lesser amounts of time which causes more strain to
themselves.

Another measurement issue is that of overlaps. In general, time use surveys fail to
include an explicit consideration of overlaps. It is seldom possible to ascertain

69

whether the time an adult spends with a child coincides with the time spent by
another adult or whether more than one child is being cared for at the same time
(Folbre et a/2005). Children may benefit from the additional time and attention they
get from more than one care pro:vider. However, the impact of child overlap is more
complex to interpret. The presence of an older child may assist the adult in caring
for the younger child and it might improve the quality of care. Whereas presence of
two young children receiving care is more demanding and stressful and hence might
result in lower quality of care.
5.3. Exploration of the association between maternal work and ehildcare Primary data
The following tables provide a picture of the relationship between maternal work and

..

child care and health that emerge out of quantitative primary data. The activities
women perform in the childcare regime, the amount of type the spend, duration and
frequency of breast-feeding, frequency of changing dress and bathing children,
Incidence of minor accidents and incidence of morbidity among children are
discussed.

70

Table 5.1: Percentage distribution of activities women perform in the childcare

fe2 ime
DW+SHG
DW+SHG
Activities in the
DW+market
DW+SHG+
activities
care regime
+micro
labour+SHG
market labour
enterprises
+micro

Feeding
Cooking food for
children
Bathing
Washing clothes
Changing clothes
Playing
Carrvinl!l'holding
Attending crying
children
Special cleaning of
the oremises
Taking to
healthcare centers
Providing special
care when sick
Extra care on
rsonal hygiene
Spending time to
enhance verbal
abilities
Assisting children
in studies
Make children
sleep
Over all
supervision
Total (Mean)

P,

enterprises
F
P
N

90
74

8
18

2
8

71
68

23
20

6
12

64

62

26
18

10
20

51
70

84
74
71
57
82
85

12
26
29
43
18
15

4
0
0
0
0
0

75
78
65
60
79
76

19
20
35
40
21
24

6
2
0
0
0
0

70
63
60
62
57
52

24
33
40
48
43
48

6
4
0
0
0
0

71
69
54
59
48
45

20
27
46
41
52
55

9
4
0
0
0
0

78

22

75

25

60

28

12

51

41

89

18

87

13

78

22

81

19

92

8,

90

10

87

13

88

12

100

100

100

100

80

20

71

29

78

22

69

31

74

26

61

39

68

27

58

29

13

91

93

85

11

82

12

100

100

100

100

37
16

12
14

83

17

78

20

72

25

69

27

F refers to fully performed by mothers, P refers to partially performed and N refers

to~ot performed by mothers

71

The above table 5.1 indicates that as women combine more activities on the work
front, the nature of their participation in childcare changes from 'fully performed' to
'partially performed'. Yet, it is noticeable that women almost exclusively perform
certain activities in the caring regime such as feeding children, putting them to bed,
helping in studies, special care when they are sick etc. irrespective of the nature of
their work.
Table 5.2: Time spent on childcare
Typology of work
A verage time spent for
children in the age gro~ 0-3
DW+SHG
8.21
DW+SHG+ME
6.34
DW+ML+ SHG
5.25
DW+ML+SHG+ME
4.87

Average time spent for

children in the age group 4-7
4.52
4.4
3.32
2.72

Average time per day (not taking into account the overlaps)
The literature suggests that the reason for the adverse impact of women's work on
child well-being is the lack of time at their disposal for childcare. The above table
shows that the average amount of time spent decreases when the number of workrelated activities of women increases. As expected, the time spent for children in the
age group of 0-3 is higher than that for children in the age group 4-7. Though the
total amount of time spent on childcare by the mothers work outside home is less,
they perform the majority of activities in the care regime as revealed in the table 5.2.
It indicates that the intensity of caring work is high when women combine more
aqtivities with regard to work outside home.

The indicators of childcare considered for the analysis using primary field data are
the following.

72

I. Time spent on childcare

2. Frequency of changing clothes and bathing
3. Median duration of breast-feeding and frequency of feeding

4. Incidences of minor accidents such as cuts, wounds, bums etc.

5. Incidence of morbidity (reference period is the two previous months for
specific illnesses such as .diarrhea, fever and cough, and three months for any
illness).
Table 5.3: Median duration an dfrequency 0 fb reast- eedin2
Median duration of feeding
Work Typology
Frequency of feeding
(months)
in a day (times)
20.18
DW+SHG
7.50
DW+SHG+ME
18.89
6.13
22.24
DW+ML+ SHG
3.43
22.88
DW+ML+ SHG+ME
2.97
The data on children in the age group of 2-7 has been used for the above table. It
shows that the median duration of feeding is high among women who perform more

..

number of activities on the work front whereas the frequency of feeding in a day is
less for them. This high figure for duration of feeding is explained by women in the
focus group discussion as a means to compensate for the less frequent daily feeding.
T a bl e 5.4: Frequenq of chan2ing c Ioth es an dbath'mg_

Work typology

DW+SHG
DW+SHG+ME
DW+ML+SHG
DW+ML+
SHG+ME

Children in the 0-3 age group Children in the 4-7 age grou
Bathing or
Changing
Changing
Bathing or
cleaning
clothes
cleaning
clothes (Avg.
no)
5.34
2.30
1.34
3.40
4.89
2.47
1.45
3.89
2.31
5.88
1.23
4.21
5.93
2.14
1.44
4.:p

73

The frequency of changing clothes and bathing children is taken as an indicator of

childcare. It is found in the above table that in the age group of 0-3, the frequency of
changing clothes and bathing is high among children whose mothers engage in
market labour and micro enterprises activities in comparison to children whose are
not so engaged. Whereas such a pattern is not found among the children in the age
group of 4-7. Children in the age group of 0-3, whose mothers work away from
home and perform more economic activities, are taken care of by substitute childcare
providers. Since the substitute care provider is less engaged in other responsibilities
while caring for the children unlike the mothers, they can pay more attention to
childcare. Moreover, mothers constantly monitor the substitute care resulting in
better care given by the substitute care provider.

Figure

5.~: percentage of minor accidents by-~~-~~

DW+ML+ SHG+ME

DW+ML+SHG

DW+SHG+ME

DW+SHG

10

15

20

25

30

Pen:entage

Data on the incidence of minor accidents including cuts, wounds and bums during a
reference period of the previou~ six months i~. collected and the responses are plotted
in the chart above. It is interesting to see that children of women with fewer work-

74

related activities, who are not engaged in income generation work either through
market labour or micro-enterprises, have greater incidence of minor accidents. This
aspect has been probed in the qualitative round and it is mentioned by some
respondents that, in general, women go out for work when more childcare providers
are available at home. Unlike mothers, substitute care providers are left with lesser
amount of other household responsibilities. Thus the multiplicity as well as intensity
of work is lower in their case. Moreover, it is the responsibility of the substitute care
provider to handover the chiia to the mother without any injuries or else it
considered negligence. Whereas when mothers themselves attend to their children,
this is not the case.

In both the above indicators of childcare, it has been found that substitute care is
better in terms of providing quality childcare. This is in corroboration with the
observation by Desai and Jain (1995) in rural Karnataka that maternal care may not
always be of better quality. Though they may not be working, mothers may have
additional burdens that constrain their involvement in childcare.

Table 5.5: Percentage of morbidity incidence bywork typology

Work typology
Incidence of
Incident of
Incidence
cough
of fever
diarrhea
DW+SHG
12
23
29
24
19
DW+SHG+ME
9
24
34
\3
DW+ML+SHG
27
30
16
DW+ML+SHG+ME

Any
illness
41
39
46
45

file above table shows that children of mothers working outside home, viz. involved
in market labour as well as micro enterprise activities do show slightly higher
incidence of morbidity episodes. However, whether maternal work is a significant

75

factor impacting adversely on morbidity incidence among children is assessed

through the analysis provided in the next section.
5.4. Analysis of factors determining maternal time spent on cbildcare - primary
data
It has been widely discussed in the literature that the amount of time mothers spend
on childcare activities is significantly influenced by their work status. Hence, in
order to examine the factors that influence time spent on childcare, an analysis has
been carried out as follows. Each activity of the child during a full day has been
recorded along with the person attending on it and the amount of time spent on each
activity. In the case of more than one child in a household, the youngest child is
selected. The amount oftime mothers spent on childcare is calculated from this data.
An ordinary least square (OLS) regression model has been used to quantify the effect
of independent variables on maternal time spent for childcare. The response
variables selected are the type of maternal work, whether living in colonies, sex of
the child, presence of a grandmother in the household, presence of elder sibling(s),
and average number of days of e<mployment of the father in a week.

76

Table 5.6: Description of variable selected for the analysis of time spent on
childcare
Description
Mean
Variable
In completed years of schooling
Education level of mother
7.2
DW+SHG
0.43
Maternal work status
DW+SHG+ME
0.15
DW+ML+SHG
0.25
DW+ML+SHG+ME
0.17
Yes =1
Living in colonies
0.23
No=O
Male -I
Sex of the child
0.52
Female = 0
Yes-I
Presence of grandmother
0.20
No=O
Presence of elder sibling(s) Yes = 1
0.34
Na=O
Average work days of Average days in a week
2.90
father

.
Ta ble 57
. : R{egressIOn resu ts 0 fd etermmants 0 f tIme
spent on c h'd
II care
Independent variables
Maternal education
Maternal work
DW+SHG
DW+SHG+ME
DW+ML+SHG
DW+ML+SHG+ME
Living in colonies
Sex of the child
Presence of grandmother
Presence of elder sibling
Average work days of father

Coefficients
0.6328 {0.2126)

-0.4982 (0.1308)
-0.7284 (0.1123)***
-1.9139 (0000)***
-0.3456 (0000)***
-0.3648 (0.2849)
-0.1842(0.04521***
-0.0421 (0.2849)
-0.0421 (0.7852)

Constant

6.3478

R square

0.4108

Number of women

421

~ote: figures in parenthesis are 'p' values

77

5.4.1. Discussion
It is found that women handling multiple tasks such as domestic work, market labour

and engaged in micro enterprise production spend significantly less amounts of time
on childcare activities. However, whether these multiple tasks of mothers adversely
affect child health is assessed in the following analysis.

Living in SC and ST

colonies considerably reduces the maternal time spent on childcare. In focus group
discussions, women living in colonies mention that neighbouring families take up
childcare responsibilities when they go for work. This might be due to (a) the
community life in the colonies,.(b) the same caste affiliation of the families living in
the colonies. This indicates the significance of social networks in childcare. The
presence of a grandmother in the household also reduces the time spent by the
mother on childcare.
5.5. Analysis of the impact of women's work on child health - Primary data
In the present analysis, an attempt has been made to understand the role of the
mother's work status in determining child health. For this purpose, the incidence of
any illness of the child during the previous month is selected as the dependent
variable. Since the dependent variable is dichotomous in nature, logistic regression
has been adopted to quantify the net effect. The type of work women are engaged in,
their educational status, their caste, whether living in colonies, sex of the child,
Jresence of a grandmother in the household, presence of elder sibling(s) and the
index of female autonomy constructed, have been selected as the explanatory.
variables.

,
\

78

The female autonomy index constructed using primary field data measures the
autonomy of women both in

th~.public

and private domain using relevant indicators

identified by women in the focus group discussion and finalized in consultation with
key informants identified, A detailed account of variables included in the indices of
autonomy in both public and private domains is provided in chapter seven,
Table 5.8: Description of the explanatory variables selected for the analysis of
f actors d etermID1D2 IDCI
. .d ence 0 fany
'lI
I ness among Ch'ld
I ren
Explanatory variables
Value
Frequency
Caste
Scheduled Caste
129
0
Others
461
I
Sex of the Child
Female
0
283
Male
I
307
Education
Illiterate
50
0
Primary
I
132
Middle school complete
234
2
High school and above
3
174
Living in colonies
Yes
0
136
No
454
I
Presence of grandmother
Yes
188
0
No
I
402
Presence of elder siblings
Yes
201
0
No
I
389
Maternal work status
DW+SHG
0
254
DW+SHG+ME
I
lOS
DW+ML+SHG
2
130
DW+ML+SHG+ME
3
101
Female Autonomy-Public sphere

>~

2-4
<4

320
213
57

0
1

481
109

Female Autonomy-Private sphere

0-2
<2

79

Table 5.9: Incidence of any illness among children on selected explanatory

variables - Odds ratios
Dependent
Explanatory variables
variable:lncidence of
any illness
Caste

SC/STrc
Others
Sex of the Child
Male rc
Female
Education
Illiterate rc
Primary
Middle school complete
High school and above
Living in colonies
Yes rc
No
Presence of grandmother
Yes rc
No
Presence of elder sibling
Yes rc
No
Mothers work status
DW+SHG
DW+SHG+ME
DW+ML+SHG
DW+ML+SHG+ME
Female Autonomy-public sphere
>2 rc
2-4
<4

0.90**

0.92

0.89**
0.86
0.79

0.82***

1.18

0.98

1.07
1.32**
1.08
0.94
0.91

Female Autonomy-private sphere

0-2 rc

<2

0.99

-2loghkehhood

78.50

cbnstant

2.46

*p<.01; **p<.05; ***p<.IO

\

80

5.5.1. Discussion
It is found that compared to the children of illiterate mothers, children of mothers
with at least primary education.llre less likely to fall ill. Whereas such a relationship
is not found when the level of education increases. Also, it is found that children of
mothers involved in the market labour in addition to domestic work have more
chance of getting affected by diseases compared to the children of mothers doing
other categories of work. Children of those who belong to the non-scheduled caste
and tribe communities and living outside the housing colonies have significantly less
chances of being affected by any illnesses. All the other explanatory variables
including female autonomy do not show any significant influence on incidence of
child morbidity.
The study sample includes the SC/ST population who have less educational
achievements, are mostly involved in market labour activities such as bamboo and

reed work and live in housing colonies. Hence, their children do have a significantly
higher chance of being affected by illness. It also shows that even among the poor,
there are sections that are further marginalised in terms of their caste and other social
background, and thus face greater deprivation than others.

5.6. Qualitative data on work and childcare

~omen in general see childcare as a significant aspect of their life. The cultural

notions associated with motherhood and caring responsibilities of women are

prevalent in the selected villages as well. However, it has also been divulged in focus
group discussions that occupational aspirations' are quite high among women. The

81

search for jobs that give better social status widely prevails among women belonging
to

the lower economic strata. I Some studies have cited the reason for the same as,

apart from greater earning and more support for the family, it is seen as a means of

upward social mobility in mode~ Kerala society. Hence it can be seen as a coupling
of traditional notions of motherhood with the search for income-generating work that
confers a higher social status.
Though the fall in fertility rates have led to a reduction in the total number of
children born, on an average every woman bears and rears one or two children in her
reproductive career. This fewer number of children per woman has resulted in a high
degree of childcare demand within the family. Women have expressed the aspect of
'quality upbringing' as an important aspect of childcare. The quality upbringing
primarily refers to providing good food and English medium education to children.
However, the idea of good food includes both nutritious food as well as bakery items
that are often sweet.
Feeding children is considered entirely a mother's responsibility since it is being
seen as an extension of breastfeeding. It has been pointed out that children are
exclusively breast-fed for a minimum of the initial three months and substitute food
is given to infants when mothers go for work outside the home. It is also mentioned
in focus group discussions that the time input required in the initial years of child

~ng is substantially high and 'the degree of care substitution by other members of
the family is lower. Cooking special meals and feeding the child is a time-consuming
\
I

See, for an interesting case study ofthe vicissitudes of upward mobility and status aspiration through
the 20'" century, Filippo Osella and Caroline Osella (2000),

82

task for women, especially in the initial years after childbirth that are also seen
exclusively a mother's portfolio. Bathing, cleaning children, changing their clothes
and washing them etc. are also other tedious activities during the initial years.
Cleaning the household premises and maintaining personal hygiene, particularly
during breast-feeding are

ano~er

array of activities. Taking children to medical

centers when they are ill and providing special care is also considered an exclusive
maternal responsibility. Developing the verbal ability of children by talking to them,
especially when children are small and assisting them in studies etc. are activities of
the mother. Overall supervision of all caring activities, though they may not be
performed only by the mother, is another responsible activity in the care regime.

5.7. Summary

In the present chapter, using both primary quantitative and qualitative data, the

specificities of childcare regime are explored. Specific activities in the care regime

are listed by women in focus group discussions. It is found that though the amount
of time spent on childcare by mothers who combine more activities on the work front
is lower, they carry out the majority of the tasks involved in the care regime. Also, it
is found that in the case of two selected indicators of childcare, viz. frequency of
changing clothes and cleaning of children, it is observed that substitute care is better
in terms of providing quality childcare. It questions the assumption of the superior
qfality of maternal care.

Though not working outside, such mothers may have

additional burdens that constrain their involvement with childcare.

Quantitative

analysis shows that children of those who belong to the scheduled castes and tribes
community, living in SC/ST housing colonies and involved in market labour have

83

significantly high chances of getting affected by any illness. The study sample
includes

the SC/ST population who have fewer educational achievements, are

mostly involved in market labour activities, such as bamboo and reed work, and live
in housing colonies. Hence, their children do have a significantly higher chance of

being ill. It also shows that even among the poor, the further marginalised sections in
terms of their caste and other

~ial

background, face more deprivation than others.

Qualitative data primarily show that child caring responsibilities place a major
constraint over women who work outside their homes and combine many jobs.
However, it does not mean that this constraint effectively prevents women from
taking up work-related activities. The constant effort and struggle to maintain a
balance between the two has been observed both in the focus group discussions and
in the case studies analysed.

\
\

84

Cbapter6

Exploring the Relationship between Women's Work and

Female Autonomy -Secondary Data (NFHS II)

6.1. Introduction
Female autonomy has widely been used in the demographic literature as a factor that
contributes to reduction in child mortality and fertility. The two important proxies of
female autonomy, widely accepted in the literature, are female education and work
status. The analytical framework that examines the impact of women's work on child
health employs female autonomy as an intermediate variable.

The underlying

assumption is that employment outside home might help women to make more
autonomous decisions that are favourable to children which in tum improves their
health in varied ways.
Quite contrary to our expectati.qn, it has been found in the analysis carried out in
chapter four using NFHS II data that female autonomy significantly enhances the
chances of morbidity in children. In this context, it is important to understand the
factors determining female autonomy as measured in NFHS II. An analysis in this
regard has been conducted.
6.2. Factors determining autonomy - NFHS II data

It il expected that education, work participation and exposure to mass media are
.some of the means by which women gain status and autonomy, both of which are
'.

important for their empowerment (Report ofNFHS II, Kerala, 2001). Hence in order

85

to measure their status and autonomy, NFHS II asked questions on women's

participation in household decision-making, their freedom of movement and access
to money that they can spend.
to have

It is also anticipated that women's autonomy is likely

significant impact on the demographic and health seeking behaviour of

couples by altering women's relative control over fertility and contraceptive use, and
by influencing their attitudes and abilities that might enhance the quality of health
outcomes of themselves and their children (Sen and Batliwala, 1997).
ki OK ID
. dIcatOrs- NFHSII
..
TabIe 61
: Houseb0 Id d eclSloo-ma
Others in
Respondent
Type of
Respondent Husband
the
with
Decisions
only
only
household
husband
only
What
items to
56.6
3.1
5.7
15.9
cook
Obtaining
healthcare
54.6
20.0
13.3
7.4
for herself
Purchasing
jewellery
or other
20.7
30.9
10.5
25i.l
major
household
items
Going and
staying
with her
20.5
29.9
10.4
31.7
parents or
siblings
How the
money she
30.3
1.8
42.3
21.4
SWill
used 16

Respondent
with others in
household

Total
percent

18.6

100

4.7

100

11.7

100

7.6

100

18.6

100

(Source: NFHS Report, Kerala, 2001)

. cash
or women earning

16 p

86

The above table provides a picture of the major decision makers of key decisions in
women's lives identified in the NFHS II. It shows that around 55 percent of women
make independent decisions on what to cook and matters regarding their own health
care. Since cooking at home is a part of social reproduction and a female gendered
activity, high rates of decision-making in this head cannot be seen as an indicator of
their autonomy.

It is also a well-documented fact that within the household,

women's health needs get less priority over that of children and other male members.
It often results in neglect of illp.ess causing a multitude of adverse impacts. Hence,
the point here is that the decision-making in the hands of women on their health
matters may not always be an indicator of their autonomy, instead it may be due to
the low level of importance accorded to it in the household. It shows that only 20
percent of women make independent decisions on major purchases including jewelry
and decisions regarding staying with the natal family. In both the matters, for around
40 percent of women, the decisions are taken without their involvement. For those
who earn income in cash, 42 percent make independent decisions on its spending
pattern. It is expected that thes.e decisions will be in favour of child well-being in
varied manner.

87

Figure 6.1: Percentage of women participating In-dlffere-n-t--household decisions and access to money

-1

Access to money

Going and staying with parents/siblings

Purchase jewellerylhousehold items

Own heallhcare

What to cook

10

20

30

40

50

60

70

80

90

Percent

(Source: NFHS II Report, Kerals, 2001)

The above chart illustrates the percentage of women involved in decision-making on
identified household and personal matters either as sole decision makers or jointly
with other family members.

88

Table 6.2: Women's autonomy bv backl!round characteristics - NFHS II

Background
characteristics

Age
15-19
20-24
25-29
30-34
35-39
40-44
45-49
Residence
Urban
Rural

Pereent
not
involve
d in any
decision
making

Percent involved in decisionmaking on

What to
cook

;tium
High
Total

Purchase
jewellery
Ihh item.

Staying
with
parents!
siblings

Go to
market

Visit
fuds
or
Rela

Percent
with
access
to
money

tive

27.8
16.9
8.9
4.6
4.5
3.0
2.8

39.4
59.1
75.6
82.4
89.5
91.8
92.8

56.2
63.3
69.0
75.1
74.9
77.6
78.0

48.7
50.7
56.7
65.2
71.5
68.1
69.6

36.9
48.9
51.1
63.7
65.8
64.7
68.4

27.5
33.6
39.1
48.5
52.7
59.4
57.1

14.5
24.9
28.9
38.3
41.8
48.4
50.7

50.3
62.4
65.4
68.8
66.4
66.8
69.9

6.8
7.3

80.7
81.0

72.6
72.5

63.9
63.2

60.3
59.6

57.7
44.7

44.6
35.8

72
64.4

4.5
6.2
9.3
7.9

88.2
84.0
76.6

78.9
72.2
65.9
73.7

69.9
62.3
56.7
65.0

66.1
60.0
54.5
59.8

53.1
47.9
48.8
45.4

44.2
39.3
37.8
34.8

56.2
58.8
63.0
76.1

5.8
10.8
4.6

82.7
75.1
87.2

72.6
73.5
70.4

65.5
57.6
68.2

62.2
62.6

55.7
34.6
48.9

43.0
30.7
35.8

72.6
52.6
73.3

5.3
(3.5)
8.1
6.9

82.9
(80.2)
78.2
83.1

69.2
(76.6)
70.3
75.1

63.2
(65.3)
58.5
67.8

59.2
(67.3)
55.5
63.6

58.3
(59.3)
49.0
44.3

45.8
38.7
37.5
36.7

67.4
(66.5)
63.6
68.3

Education
Illiterate
Literate<middl
eshl
Middle school
High school
abv
Religion
Hindu
Christian
Muslim
Caste
SC
ST
OBC
Other
Cash
employment
Work for cash
Work not for
cash
Not worked
last 12mth
Standard of
living
index

Own
health
care

Percent do not need

pennission to

8004

5404

3.0
5.6
8.5

87.7
89.7
78.6-.

76.2
67.2
71.7

74.4
67.9
59.9

72.0
59.5
56.1

62.1
44.7
43.6

50.6
35.0
34.2

76.5
74.7
62.8

6.7
7.4
7.1

82.6
79.9
81.9

71.2
72.9

7p

64.8
61.9
65.3

62.2
58.0
61.8

55.4
44.8
49.2

44.4
34.9
40.0

62.8
60.0
79.5

7.2

80.9

72.6

63.4

59.7

47.7

37.9

66.2 .

(Source: NFHS II Report, Kerala)

89

The above table shows that there is a positive correlation between age and all the
indicators of autonomy. The higher the age greater the percentage of women taking
decisions. The residence pattenI' does not show any noticeable impact on decisionmaking and freedom of mobility indicators. Except in the case of access to money,
women with education above high school are relatively less active in getting
involved in various household decision-making and have less freedom of mobility. It
is found that the share of Christian women in all decision-making as well as mobility
indicators is less in relation to Hindu and Muslim women. More women belonging to
forward castes do have decision-making power when compared to SC and OBC
women. However, in the case of mobility, particularly in public places such as
markets and also in visiting friends and relatives, women belonging to forward caste
are more restricted.

Those who work for cash are more involved in household

decision-making and have more freedom of mobility and access to cash. Women
from families with low and high living standards are more engaged in decisionmaking in relation to those having medium living standards. Nonetheless, mobility is
greater for women in the low standard of living category compared to the other two.

90

'hbea'
h . I VIO
. Iene e-NFHS n
Table 6.3: W omen s expenence Wit
bngs or PbYSIC8
Percentage beaten on
Background characteristics
physically mistreated
since age 15
Religion
Hindu
11.1
Muslim
9.1
Christian
9.7
Caste
Scheduled caste
15.3
Scheduled tribe
(\3.5)
Other backward caste
8.5
Others 17
10.8
Household type
Nuclear
12.8
Extended
8.4
Education
Illiterate
18.4
Literate abv. middle school
12.4
Middle school complete
9.4
High school and above
6.4
Age
15-19
6.7
20-29
7.0
30-39
11.1
40-49
\3.1
Standard of Living Index
Low
20.8
Medium
10.4
High
4.4
Motbers work status
Working for cash
15.3
Working not for cash
(13.5)
Not worked in the past 12 months
8.5

The table 6.3 shows the incid.t<nce of domestic violence by selected background
characteristics. A living environment without fear and violence is one of the
im\,ortant indicators of women's status in any society. Women who belongs to

.SC/ST community, relatively less educated, living in nuclear households with low
\

,
"Not belonging to a scheduled caste, a scheduled tribe or aft other backward class

91

standard of living and working are experiencing high incidence of domestic violence.
It should be noted here that incidence of domestic violence is high among working
women and other indicators of female autonomy are better among them. Hence, all
the indicators may not be unidirectional and it needs to be understood properly in
any given context.
Analysis of factors determining Autonomy
In order to examine the factors determining women's autonomy, multivariate OLS
regression analysis has been carried out with autonomy index as the dependent
variable and a set of explanatory variables have been employed.
Female autonomy has been measured in the NFHS (II) in terms of household
decision-making power, mobility status and access to the resources. The decisionmaking variables are the foIlowing. (I) who decides what to cook, (2) who decides
on obtaining health care, (3) who decides to purchase jewelry, (4)who decides about
the respondent staying with family.

The options given are: (l) respondent, (2)

husband, (3) others in the household, (4) jointly with the husband, (5) jointly with
others in the household. The mobility indicators are measured in terms of: (I)
permission needed to go to market, (2) permission needed to visit relatives or friends
and the options given are yes and no. The access to resources is measured by asking
whether the respondent is allowed to set aside money and the options are given again
yei and no. Since our analysis examines the factors determining female autonomy,
I

the context in which the decision-making takes place is important. If male members
\

are not available in the household for decision1naking,

women become the prime
,

92

decision maker. In order to avoid the same, female-headed households have been
taken out from the analysis.
A composite index of autonomy is constructed by giving values ranging 0 to I for all
the indicators and a sum total value is taken.

For decision-making variables,

'respondent' is given a value of I and 'husband' and 'others in the family' are given
a value of 0 each. 'Jointly with husband' and 'jointly with others in the family' are
given 0.5 each. For the rest of the variables, a value of I is given for 'yes' and 0 for
'no'. Hence, the autonomy index ranges from 0 to 7.
Explanatory variables considered:
(I) Age of women in five-year interval age groups from 15-19 to 45 -49
(2) Education (illiterate, primary, secondary, higher)
(3) Place of residence (urban/rural)
(4) Family structure (extended/nuclear)
(5) Religion (Hindu, Christian, Muslim)
(6) Caste (scheduled caste, scheduled tribe, other backward caste, others)
(7) Standard ofliving index (low, medium, high)
(8) Work status (working, non-working)

93

. bles- A utonomy index

Table 6.4: Significance of d epen d ent varia
Coefficient
Standard error
Explanatory variables
3.990'"
0.183
Constant
-1.604'"
0.227
Age 15-19
-1.085'"
0.139
Age 20-24
-.880'"
0.126
Age 25-29
-.494'"
0.127
Age 30-34
-.234"
0.122
Age 35-39
.01
0.124
Age 40-44
. -.587'"
0.087
Religion - Muslim
-.286
0.097
Religion - Christians
.469'
Family structure
0.084
.212**
Area of residence
0.081
-.05'
0.147
Education - Primary
.03
Education - Secondary
0.145
.201
Education - High school
0.171
.095'"
0.321
Caste-ST
.105'
0.129
Caste-SC
-.163
0.074
Caste-OBC
-.04'
0.102
SLI-Medium
.141'
0.122
SLI-High
Currently working
0.084
.370'"

R2 _0.241
Adj R2 - 0.234
Note: = .05; = .01; = .001
The significance of the explana(ory variables is explained below.
The regression analysis shows that education is not very significant in explaining
women's autonomy. The caste status and the standard of living are found to be
significant in determining female autonomy. The younger age groups have shown a
very significant negative relationship with autonomy.

Muslim women do have

significantly less autonomy compared to Hindu women. Also, living in nuclear

families as well as participating in the work force significantly enhances the
a1onomy of women.

Multiple classification analysis (MeA) is performed in order to assess the gross and
\

net effects of the explanatory variables on women's autonomy. This analysis allows
,

94

us a more accurate assessment of each factor under consideration, controlling the

effect of other variables, with women's autonomy.
Table 6.5: Impact on dependent variables on female autonomy - multiple
classification analysis
Autonomy index
Explanatory variables
UnAdj
Adj
Age 15-19
2.0932
2.2944
Age 20-24
2.4887
2.5906
All.e 25-29
2.8055
2.8035
3.2617
3.1829
Age 30-34
Age 35-39
3.5178
3.4622
All.e 40-44
3.6990
3.7138
Age 45-49
3.6101
3.6358
Beta &Eta 0.320 ;0 .297***
Religion - Hindu
3.3636
3.3096
Religion - Muslim
2.9033
3.1737
Religion - Christians
3.2405
3.1108
Beta &Eta 0.141 ;0.078**
Family structure -Extended
3.1346
3.1255
I
Family Structure- Nuclear
3.4246
3.4538
I
Eta & Beta 0.119; 0.118***
Education - illiterate
3.7344
3.4395
Education - primary
3.3815
3.2019
Education - secondary
3.1188
3.1810
Education - high school
3.0572
3.1795
Beta &Eta 0.130~0.046
Caste-SC
3.5483
3.4416
Caste- ST
3.5670
3.5913
Caste-OBC
3.1278
3.1236
Caste - Others
3.2185
3.2584
Beta&Eta 0.138; 0.124**
Standard of living - Low
3.3588
3.4583
Standard of living - Medium
3.2117
3.1359
Standard of living - High
3.1153
3.1088
Beta &Eta .09; .082**
Place of residence - Rural
3.1509
3.1628
Place of residence - Urban
3.3797
3.3394
Eta &Beta .04; .038
3.0830
3.1546
\ViIrk status - Non working
. 3.3586
Work status - Working
3.5871
Beta &Eta 0.15 ; 0.118**
N-2242
Grand Mean - 3.2124

Note:

* = .05;

** = .01 ;

*.. = .001
95

6.2.1. Discussion
All the factors under consideration have explained 23.4 percent of variation in
autonomy and a large chunk of variations are explained by factors that are not
included in the model. Even though education has shown a negative and moderately
significant relationship in the primary education category in the regression analysis
(gross effect of eta = 0.13), its net effect has reduced considerably (beta = .046) after
making allowances for other explanatory variables, indicating that education does
not have a significant bearing on women's autonomy. In the case of work status, eta
is 0.1 5 and beta has reduced to O. Jl8 and is significant at .00 I level. The gross
effect of family structure remained the same after making allowances to other factors
and both are highly significant statistically indicates the independent influence of this
variables on women's autonomy The age of women do have large net effect (beta

.30) on autonomy and in the younger age groups it is negatively significant and a
threshold age group 35-39 above which the relationship has turned out to be positive.

It shows that women belonging to backward communities with low living standards
and working have more autonomy. The positive association between female
autonomy and child morbidity found in chapter four is mainly due to this reason.

6.3. Summary

It \has been found in the analysis of secondary data (NFHS II) in chapter four that
female autonomy of women is statistiCIIlly significant in negatively influencing the
child health. Hence it is attempted in the \present chapter to understand the
constituent factors of female" autonomy as rlleasured in the NFHS II. Also,

96

multivariate analysis is carried out in order to examine the major factors that
detennine female autonomy. It is found that women belonging to scheduled castes
engaging in market labour and with low standard of income have more autonomy in
decision-making. Children of these women experience more morbidity incidence in
comparison to other children.
A pertinent question that arises here is that whether the indicators of autonomy
selected and measured in NFHS II such as decisions on what to cook, major
purchases, expending money and a range of freedom of mobility indicators, really
reflect the autonomy of women? Moreover, what exactly is the notion of female
autonomy that is being used in the conceptual framework that explains the
relationship between maternal work and child health? These questions are addressed
in the following chapter.

97

Chapter 7

Female Autonomy, Women's Work and Child Well-beingPrimary Data

7.1. Introduction
Demographic research traditionally looks at various determinants of demographic
processes such as fertility, mortality, migration etc. The relative significance and
various pathways of influence of women's education and employment in attaining
better demographic features, particularly in the context of developing countries have
been well documented in the demographic literature. Hence, these two major factors
along with other indicators such as the family structure, age at marriage, spousal age
differences etc. are constituted as a summary indicator of women's status, a
significant explanatory variable of demographic outcomes. The status of women,
often an outcome of the exercise of their individual will, particularly within the
domestic domain, is expected to be determined by these individual level variables.
Hence, these variables have largely been used as proxy variables of women's
autonomy in the demographic research. Recently, however, there has been a shift
from using proxy variables to direct measures of autonomy in the demographic
studies. This shift is mainly due to the realisation that the extent to which women
enjoy autonomy is significantly shaped by social institutions of gender within a
cfmmunity (Mason 1994; Dyson and Moore 1983; Basu 1992; Jejeebhoy 2000) and
therefore similar achievements in education or employment status might not result in
the same degree of autonomy for women belollging to different regions, religions,
castes, economic status etc. There are arguments that question the 'autonomy'

98

..

involved in such decision-making processes (Sen 20(0). Thus, it is pertinent to

understand the concept of female autonomy and need to examine the appropriateness
of using the concept in the relationship between maternal employment and child
health as an intermediate variable.
In the present chapter, an attempt has been made to examine how the concept of
autonomy is employed in the demographic literature in determining child health and
care, as one of the significant demographic outcomes. Here, autonomy is being
conceptualised as an effect of female work participation and a cause for better child
well-being. Following is a dis\)ussion on the relevance of female autonomy as an
intermediate variable in this interaction between women's work and child wellbeing.

Two major issues that emerge out of the discussion on conceptualising

female autonomy as an intermediate variable in the relationship between women's

work and child well-being are: (\) the conceptual understanding and the
operationalisation of female autonomy, (2) the appropriateness of employing the
concept of female autonomy as a determinant of child well-being. Using primary
data, two indices of autonomy are constructed, selecting key indicators of their
autonomy pertaining to their public and private realms, which are contextually
significant. The association between women's work and these two indices is
examined and the impact of such an autonomy on child health is also analysed.
7\2. Female autonomy and child well-being
In the conceptual framework that explains the possible pathways of influence
between maternal work and status on child health and well-being, female autonomy

99

has been used as an intermediate variable. It is expected that women's employment

may translate into a greater control of them over the expending of resources,
increased exposure and access to relevant information about child-bearing and
rearing practices and an enhanced ability to manipulate and engage the world outside
the home to meet the nutritive, medical and survival needs of children better. Hence,
in order to capture these effects of maternal work on child health and well-being, the
concept of female autonomy has been considered. Here, the two important questions
that need to be addressed are: (I) how the concept of autonomy is being understood
and operationalised? (2) Appropriateness of employing women's autonomy as a
causal factor in determining chitd well-being.
7.2.1. Conceptualisation oUemale autonomy and women's empowerment
As discussed in the introduction, female autonomy has been employed as a
determinant of demographic processes such as transitions in fertility and mortality
and it primarily attempts to capture the status of women in a given context. Since
female autonomy is a multi-dimensional entity, which refers to different aspects of
women's life, it has variously been defined. Dixon (1978) has defined it as 'the
degree of women's access to, jIIId control over, material resources (food, income,
land and another forms of wealth) and to social resources (including food, income,
power and prestige) within the family, in the community, and in the society at large'.
'J\he well-quoted study of Dyson and Moore in the Indian context on 'On kinship
structures and female autonomy' (1983) defines autonomy as 'the capacity to
manipulate one's personal environment and' the ability

technical, social and

100

psychological to obtain information and to use it as the basis for making decisions
about one's private concerns and those of one's intimates'.
The concept of autonomy is more static in nature and it refers to women's status at a
given point of time. Whereas the more popularly used indicator of female status, i.e.
empowerment conceptualised both as a dynamic process and a goal, has received
more attention in the demographic research particularly after its recognition and
pervasiveness in the International Conference in Population and Development
(ICPD) document as a precondition for achieving better reproductive health and
rights for women. Empowerment is defined as a process by which the powerless gain
greater control over the circum,stances of their lives. It includes both control over
ideology, beliefs, values and attitudes (Batliwala 1994). It means not only greater
extrinsic control, but also a greater intrinsic capability - greater self-confidence, and
an inner transformation of one's consciousness that enables one to overcome external
barriers to accessing resources or changing traditional ideology (Sen and Batliwala
2000). Sen and Batliwala have identified th~ four levels at which the unequal power
relation operates. They are (I) the household/family, (2) the community/society, (3)
the market, (4) the state. These different levels of power relations operate as a
closely woven mesh of power, implying that the unequal relation at one level may
get reinforced by the same at another level. Also, even if the relations are eased at
one level, it may continue to hold women at another levels.
Paula England (2000) argues that the objective bases of power (economic resources,
laws, norms, and rules), combine interactively with subjective states of entitlement
and efficacy leading to an exercise of power, and this leads to outcomes in women's
101

self interest. By using the term 'self interest', she delineates the notion of 'altruism'
from it that often drives the individual action of women in comparison with that of
men.
From the above definitions and interpretations of empowerment, it is evident that
empowerment is a more powerful and dynamic concept in understanding women's
subordinated social position as against autonomy. However, the operationalisation of
the concept in a given context stands as a major challenge.
7.2.2. Operationalising tbe concept by employing appropriate measurement
The conceptualisation of empowerment as a process and an outcome, gaining control
over various material resourcc;s as well as

ideological bases of gender, poses

challenges to the operationalisation of the concept. It certainly requires in depth

qualitative explorations.

Demographic research, essentially

known

for

its

quantification of social factors, largely uses female autonomy as an indicator of

women's status particularly due to its. complaisance towards quantification.
However, there are certain serious drawbacks in its operationalisation by the
commonly employed indicators for measuring female autonomy.

These are

discussed below.
(1) Most of the commonly used indicators of female autonomy are employed to
measure the relative status of women within the household. It does not take into
account the larger context, either dommunity or region, within which the family
as a microcosmic cite of gender power relations operates. If the community or
society is characterised by high levels pf domestic and social violence,

102

widespread dowry practices, high alcohol consumption by men and relatively

less share of male income goes for household maintenance, its interaction with
household level variables and the resultant outcomes in terms of autonomy needs
to be examined.
(2) The specific indicators of autonomy are derived from a typical patriarchal family
set-up wherein all the decisions are taken solely by men and female mobility is
highly restricted. The changes that take place in the society and economy across
time transform the gender relations within the context of family. Hence, many of
the measurers of autonomy are unable to capture the changed dimension of the
gender power relations within the family. Since it assumes a typical patriarchal
family in which men take all the decisions, the highest level of autonomy is
identified in cases where women are the sole decision makers, which is erroneous
due to two reasons. Often, the household management and decision regarding
the same lies with women in families experiencing severe resource constraints. In
female-headed households too they are .the sole decision makers. In these two
contexts, the decision making of women cannot be viewed as autonomy or high
status, rather it shows the relative weaker position of women. Again,
empowerment of women within the family context envisages equal relations of
power between men and women and the values of democracy being cherished.
Thus decision-making should be a consultative process rather than that of a
single individual.

(3) Yet another issue is that of the nature of 'decision-making that is employed to
measure autonomy. Mobility indicators measure by and large the freedom of

103

mobility to attend social functions, marriages, going to temple etc. The majority
of household decision-making measures are with regard to the management of
the household budget, purchasing of small consumables, children's education and
health, inviting guests and giving gifts etc. All these decisions come under the
umbrella of 'social reproduction', which is culturally a female gender activity in
the division of labour.

Thus, they cannot be viewed as indicators of

empowerment and higher social status of women, rather they indicate the
reinforcing of gendered labour division.
(4) Another related question is that of the basis of women's decision-making in
terms of individualism or altruism. Sen (1990) in his exposition of a household
where gender relations are characterised both by co-operation and conflict, points
out that since women's perceived interest is so intimately linked to the family's
welfare, it could influence bargaining outcomes such as to make a perceived
interest choice, weakening their well-being. Due to this reason, the decisionmaking outcomes may not often be that 'a!itonomous'.
7.2.3. Employing the concept of women's autonomy in child well-being
In demographic research, female work status is being considered as a factor
promoting better child health and well-being through their enhanced autonomy. The
ways in which the economic activity of women through their independent decision'Ildtong result in better child

h~alth statps has already been explained earlier. Here

the question that needs to be addressed is whether equipping women to take care of
their children in a better way (either through sp~nding their income for good care

104

services, providing nutritive food etc. or through acquiring knowledge on better care
practices) is the content of female autonomy? Whether these decisions are quite
independent of the ideology and institution of motherhood? If they are not
autonomous from the cultural notion of the institution of motherhood, they will not
lead to the empowerment of women, instead they reproduce the institution of
motherhood, which is one of the key institutions that systematically maintains gender
inequality.

7.3. Women's autonomy and work status - Primary data

In the context of the above discussion on conceptualisation and operationalisation of

women's autonomy/empowerment and the appropriateness of employing commonly

measured indicators of female autonomy as a factor promoting child well-being, in
the present chapter, we attempt to explore the relationship between women's work
and female autonomy and the aSsociation between female autonomy and child health
status. Since the commonly used measures of autonomy have certain limitations,
two indices of autonomy have been constructed. One index includes indicators of
women's visibility in the public sphere and the other index includes variables that
capture their relative position in the private domain. The relationship between work
and these two indices of autonomy has been explored.

Using primary data, two indices of autonomy (in the public and private spheres) are
Jnstructed in order to understand whether a similar relationship to both the indices

.exists with work status of women and child health. The relevant questions are
selected by discussing with both women me1pbers in the self-help groups and

105

identified female key infonnants in the locality, such as the previous women
panchayat president, a political activist, an active member of the local NGO and a
female activist from the Kerala Shasthra Sahithya Parishad (KSSP)18. It is important
to understand what women mean by autonomy in the given social context. Therefore,
initially a list of key indicators of their social status including decision-making that
they felt to have significant bearing on their social position in the public as well as in

the private domain were listed. Then women were asked to prioritise them. Among

the list of prioritized key indicators, six indicators were selected pertaining to each
public and private realms in consultation with key infonnants.
The indicators of autonomy in the public sphere are the following
I. Took part in active discussion on development needs in gramsabha l9? (yes =1, no
=0)

2. Participated in the preparation of development projects20 for women? (yes= 1, no

=0)

..

3. Aware of jagratha samith;21 for addressing issues of violence against

women?(yes=l, no = 0)
4. Involving in other political, social or cultural activities? (yes = I, no = 0)
5. Involved in any protest for the rights of women or in general protest in your
locality? (yes = I, no = 0)
,. KSSP is voluntary people's science movemellt active in Keral.
,. Gramsabha is village assembly
. 20 Development projects are the sector wise projects prepared at the local panchayat level as part of
the annual planning process. There are' projects under WCP for exclusive women beneficiaries
21 Jagratha samithi Jagratasamiti is a body constituted in all the local governing institutions under the
aegis ofKerala State Women's Commission to address the issue of violence against women more
effectively

106

6. Do you feel free and confident to go to all major public offices in the local
panchayat? ( yes = I, no = 0)
Each of the above indicators takes a value of 0 or I and hence the total value ranges
from 0 to 6.
The questions are employed to capture the autonomy of women in the private realm
are the following.

I. Is there any active discussion with spouse on matters related to household

management and areas where you have difference of opinion? (yes = I, no = 0)
2. Ever experienced physical violence after your marriage? (yes = O,no = I)
3. Ever experienced emotional torture and mental violence after your marriage?
(yes = 0, no= I)
4. Whether the domestic work is adequately shared by your husband? (yes = I, no =

0)
5. Will you be able to take independent decisions to go to places you wish, such as
meeting friends or relatives or attending some public functions? (yes =1, no = 0)

6. Whether decisions on reproductive matters such as number of children,

contraception etc. are jointly taken? (yes =1, no = 0)

All the six indicators take a value of either 0 or 1 and the total value ranges from 0 to

6.

107

Table 7.1: Mean of autonomy indices by workJypology

Type of work
Mean Autonomy
Mean Autonomy index in
index in the public
the private sphere
sphere
DW+SHG
2.78
1.21
DW+SHG+ME
3.81
1.52
DW+ML+SHG
2.62
1.18
DW+ML+SHG+ME
3.64
1.33
Total
3.21
1.31
It is found in the above table that the autonomy women enjoy in the private realm is
considerably less compared to that in the public domain. Now in order to understand
the factors that influence this measured autonomy both in the public and the private
domains, OLS regression analysis has been carried out. To examine the independent
impact of explanatory variables, Multiple classification analysis is made use of.
Set of explanatory variables selected are the following
(I) Age of women in five year interval groups from 15-19 to 35+

(2) Education (primary, secondary, higher)

(3) Family structure (extended Inuclear)

(4) Religion (Hindu, Christian, Muslim)
(5) Caste (scheduled caste, Scheduled tribe, other backward caste, others)
(6) Work status (DW+SHG, DW+SH<.:i+MI, DW+ML+SHG, DW+ML+SHG
+ME)

108

. bles on Iiema e autonomy IDdlCes

Ta bl e 72
. . S,12nlfiIcance 0 fd epen d ent varia
Explanatory Variables
Autonomy in the public
Autonomy in the private
Coefficient
Standard
Coefficient
Standard
error
error
Constant
3.990**
0.134
5.581"
0.332
Age 15-19
-1.004***
0.219
-0.712
0.132
Age 20-24
-0.885"
0.139
-0.996**
0.079
-0.780**
0.126
AlI:e 25-29
0.064
-0.681*"
Age 30-34
0.494**
0.127
0.443*
0.221
Religion - Muslim
-0.587
0.087
-0.254
0.438
Reiill:ion - Christians
0.086
0.097
0.013
0.336
Family structure
-0.196**
0.084
0.262*
0.118
Education- Secondary
0.163
0.145
0.142
0.098
Education- High school
0.720"
0.171
0.084
0.229
Caste -SC&ST
0.095**
0.321
0.312
0.144
Caste-OBC
-0.163
0.074
0.187
0.192
DW+ME+SHG
0.761
0.\03
0.292
0.112
DW+ML+SHG
-0.376**
0.066
0.199
0.153
DW+ML+ME+SHG
0.442**
0.117
0.346
0.082

Note: p< 0.05; p< =0 .01; ... p< 0.001

7.3.1. Results of the impact of dependent variables on female autonomy indices

In the regression analysis, age is found to significantly influence women's autonomy
in the public sphere as well in the private sphere. Religion has negative association
with autonomy in the public sphere in the case of Muslim women. Extended family
structure is found to be positively significant in determining female autonomy in the
public sphere whereas it negatively influences the same in the private sphere.
Education above the high school level is statistically significant in determining
women's autonomy in the public realm whereas it is found to be insignificant in the
case of female autonomy in the private domain.

Type of work significantly

4etermines women's measured autonomy in the public sphere. Women involved in

109

SHG activities and micro-enterprises do have significantly more autonomy in the

public domain in comparison to those engaged in market labour in addition to selfhelp activities. Whereas in the private sphere, those who are engaged both in market
labour and micro enterprises are found to have moderately significant impact on
female autonomy. The following MCA table provides a better picture of the impact
of independent variables on autonomy.
Table 7.3: Influence of dependent variables on female autonomy indices multiple classi filcation analysis
Explanatory variables
Autonomy in public
Autonomy in private
UnAdj

Adj

2.33
2.22
2.56
2.48
2.87
2.78
3.32
3.30
3.41
3.36
Beta & Eta: 0.14; 0.12***
3.34
3.22
Religion - Hindu
Religion - Muslim
3.11
3.18
Religion - Christians
3.38
3.32
Beta & Eta: 0.09; 0.06**
3.99
3.97
Family structure -Extended
3.71
3.70
Family structure- Nuclear
Beta & Eta:O. II; 0.10***
3.50
Education - Primary
3.68
Education - Secondary
3.72
3.68
Education - High school
4.13
3.89
. Beta & Eta: 0.09;0. 04***
3.89
3.92
Caste - Other
3.63
Caste -SC&ST
3.61
3.84
Caste-OBC
3.90
Beta & Eta :0. 07; 0.04**
3.66
DW+SHG
3.70
4.09
4.11
DW+ME+SHG
3.31
DW+ML+SHG
3.33
3.94
3.93
DW+ML+ME+SHG
Beta & Eta: 0.12; 0.10"*
Gnmd mean - Autonomy mdex: Pubhc sphere -3.12
Age 15-19
Age 20-24
Age 25-29
Age 30-34
Age 35+

Unadj

Adj

1.50
1.51
1.72
1.68
1.77ple
1.79
1.80
1.77
2.01
1.86
Beta & Eta :0. 09; 0.07**
1.78
1.72
1.74
1.73
1.92
1.99
Beta & Eta: 0.08; 0.06
1.68
1.66
2.01
1.99
Beta & Eta:O. 07;0. 08***
1.81
1.79
1.87
1.86
1.96
1.88
Beta & Eta: 0.08;0. 04
1.78
1.70
1.99
1.90
1.86
1.82
Beta & Eta :0. 06;0. 03
1.78
1.75
1.86
1.80
1.83
1.81
1.98
1.97
Beta & Eta:O. 08;0. 03*

qrand mean - Autonomy index: Private sphere - 1.88

110

7.3.2. Discussion of the results of the impact on selected dependent variables on

female autonomy
The independent impact of age 'on female autonomy particularly that in the public
sphere is evident from the table 7.3 (Beta & Eta: 0.14; 0.12) and is highly
statistically significant. Young women below 30 years of age have significantly less
autonomy compared to older women. Muslim women have significantly less
autonomy in the public realm compared to women from the two other religious
groups. Such a significant association between religion and autonomy is however
absent in private chores. Family structure is found to have an independent influence
on autonomy, extended structure in the pubic domain (Beta & Eta:O. II; 0.10) and
nuclear structure in the private domain (Beta & Eta:O. 07;0. 08). This result has been
further probed in the qualitative round to validate the result.
The independent impact of education above the level of high school on female
autonomy in the public sphere reduces when discounting the combined influence of
other variables (Beta & Eta: 0.09;0. 04) though it is statistically significant. Women
engaged in micro enterprises & SHG activities, and those who combine these two
with market labour activities are found to have more autonomy in the public realm in
comparison to those who do only domestic work and SHG activities, as well as
.'

women who engage in domestic work, market labour and SHG activities. It shows
that engagement in micro enterprises and market labour activities combined with
general SHG activities gives women more autonomy in the public realm (Beta &
Eta: 0.12; 0.10). This result has also been further explored in the qualitative round in
order to better our understanding on social realities over the observed statistical
si~ficance. In the private domain, women who combine all the activities including
\

III

market labour and micro-enterprise production have more autonomy in comparison

to women in other work typologies. However, it is not highly significant statistically.

As discussed earlier, it is expeCted that women's autonomy plays a positive role in

enhancing child well-being. To examine the same, the two indices of autonomy are
included as independent variables in the regression analysis using primary data that
determine child morbidity, carried out in chapter five. Regression analysis showed
that female autonomy does not significantly determine child health morbidity. It does
not mean, however, that women's autonomy in the public and private chores does
not influencing children in any possible way. We need to keep in mind that our
analysis is limited, with child health measured in terms of morbidity incidence.
Women's presence and involvement in public decision-making at the local level and
their engagement in the key decision-making within the domestic domain influence
the socialization of children which might probably gear them up towards adopting
non-discriminatory gender norms and practices.

7.4. Qualitative data on female autonomy

In the focus group discussion, as in the case of NFHS data, questions were asked
about household decision-making power. Questions about each aspect of household
decision-making, viz. child healUt, small purchases at home etc., the stylized set of
questions that measure autonomy, have been asked. It has often been reported by
women that they themselves make such decisions. The discussion gave clear
indication that since women started getting income, however small it may be, men
~

slowly withdrawing from the responsibility of meeting daily household expenses.

112

Also, many women have mentioned that men's contribution to household income has
reduced after women started earning. Thus, women see it not as an indication of their
enhanced 'autonomy' but as 'neglect' on the part of men that often results in women
getting over-burdened. This re1{~als that decision-making within the household does
not mean the same to all women across different geographical locations and social
axes of power. It might have an entirely different meaning in a different social
setting, which is characteristically different. As a result, it has been attempted to
formulate an alternate index of autonomy that incorporates certain identified key
decision-making outcomes of women in the study context, pertaining to both their
public and private realms.
It shows that the measured autonomy in the public sphere is considerably higher than
that in the private realm. The qilestion of individual agency vs. collective agency is
relevant here. It might be the fact that the newly opened up state-sponsored public
space of PRis has provided some space for women in public and it could to a certain
extent effectively bring women into the public sphere through self-help activities,
whereas an equivalent corresponding shift in the private domain has not been
observed. This is because of the social context which is characterized by gender
norms and practices as observed in the qualitative round, viz. high prevalence of
dowry practices, and acceptance of wife beating as a social norm etc. by women.
Again, it may be noted that the presence in the public sphere is also highly structured
by the State. It is obvious that the presence of women is pervasive only within the
~

sponsored PRI institutional structures. From the qualitative round it is observed

that there were occasions where the collective articulation of female alZency could

113

have been possible in situations such as sexual abuse by the panchayat secretary on a
daily wage female worker, corruption of an official in the Krishi Bhavan (village
agricultural office) and interference of the left party in the identification of
beneficiaries for development projects. The Kudumbashree SHG groups have taken
up none of the above issues.
Yet another feature of the public space that SHGs provide is that it hardly challenges
existing gender norms in the society. Only women, working in groups including
both young and old age groups are in close conformity with the local gender norms.
Moreover, the flexibility of working time and the proximity with the household, in
the case of SHG activities and micro enterprises, considerably reduces the tension
between domestic responsibilities and the work outside.
As discussed earlier, one of, the major limitations in understanding women's
decision-making within the household and the community by employing
conventional questions with dichotomous set of answers as yes or no, is that it totally
leaves out the processes involved in the decision-making mechanism. These
processes, including an array of negotiations and bargaining in varying degrees, are
perhaps more important that the outcomes of the decision-making process.

To

capture such instances involving negotiations and bargaining, selected case studies
have been used.

7.5. Summary
The analysis made in the previous chapter has revealed that female autonomy

~easured in NFHS II does not really reflect the agency that women achieve through
114

their work and the autonomous decisions they make. Consequently, a critical
evaluation of the concept of autonomy used in demographic literature and the
conceptualization of autonomy as an intermediate variable in the relationship
between women's work and child well-being is made. The issues involved in the
operationalisation of such a concept are also dealt. Alternate measure of autonomy
that is contextually relevant is. constructed. They are two indices that measure
autonomy of women both in the public and the private domain. It is found that the
autonomy in the public domain is significantly determined by work status of women.
Those who engage in Kudumbashree SHG activities and micro enterprises
production are significantly having more autonomy in the public domain. Such a
relationship is not there between work status and autonomy in the private domain.
This might be due to two reasons. One the one hand, autonomy as measured using
primary data is considerably associated with the activities of the self-help groups and
local panchayat resulting in naturally a high degree of the same among who are
engage in it. Moreover, it is a highly legitimized social space and hence the
participation in it is no more a matter of dissent in the family or community. On the
other hand, the autonomy in the private domain is highly constrained by gendered
practices and norms in the family and community such as dowry practices, high
degree of violence against women, women themselves justifying the act of violence
inflicted on them etc. that effectively prevents them to translate the power and
autonomy that they gain in the public domain into the domestic sphere.

liS

Chapter 8

Childcare arrangements -Role of Formal and Non-formal

Institutions
8.1. Introduction

The type of available institutional arrangements for childcare is one of the major
factors that determine the impact of maternal employment on childcare. There are
two types of institutional arrangements, viz. formal and non-formal. The formal
institutional arrangement includes anganwadi centres, a state-owned public
institution meant for pre-school children and other private childcare centres for
infants, and kindergarten schools for pre-school children. The informal institution is
the family in which childcare is undertaken by its members in varied ways. Since
maternal care is culturally treated as the 'real' care, all other childcare arrangements
other than the same are labeled as 'substitute care'. The idea is that it basically
substitutes maternal care and hellce has the culturally loaded assumption of being of
'naturally inferior' quality to maternal care. There are many studies that evaluate the
quality of such substitute care.

In the present chapter, an assessment of the existing institutional care arrangements

and of the issues pertaining to the same is made. The NFHS I and II surveys provide
information on childcare providers when mothers go for work.

116

Table 8.1: Childcare provider when tbe mother works away from borne (0-4 age
!!roup) NFHS I
Type of childcare providers
Percentage
distribution
HusbandlPartner
2.8
Older children
5.8
Other relatives
65.3
Neighbours
2.3
Servantsihired help
7.4
Child is in school
6.5
Institutional care
9.3
Other
0.9
Total
100
When women work away from home, the major childcare providers are other
relatives and their share is 65.3 percent. It includes grandmothers and all other
relatives. Institutional care is around 10 percent and the paid servants about 7
percent. These data refer to children below the age offour years.
The female work participation rate according to NFHS I is 25.7 percent. Out of them.
17.4 percent work at home and the remaining 82 percent work away from home. As
we discussed in the chapter exploring women's work, women often combine both
productive and reproductive IaQaur. In order to understand whether they combine
childcare with income-eaming work when they engage in home-based work, the
following question has been asked in NFHS I. While you are working, do you usually
have (name of the youngest child at home) with you, sometimes have himlher with
you, or never have him/her with you? The following pie chart depicts the responses
for the same question.

117

.. - - - - - - - - - - - - - - - - -

,--------~-

Figure 8.1: Children with mothers while working

Usually
7%

Sometimes
17%

It shows that around 25 percent of women combine childcare along with their

income- earning work.

In NFHS II, the female work participation rate has reduced to 23 percent from

25percent in the first round. However, the share of women who engage in homebased income earning activities has almost doubled to 32.5 percent. As a result, the
share of women who take care of children while they work has also increased
correspondingly (See the following table 8.2).

118

Table 8.2: Care provider for children in the 0-3 age group when mothers work
OOHS II)
Type of childcare providers
Percentage
distribution
Respondent
42.9
HusbandIPartner
2.4
Other relatives
28.6
Neil!hbours
2.4
Servantslhired help
9.5
Child is in school
1.2
Institutional care
9.5
Other female child
1.2
Others
2.4
Total
100
NFHS II does not provide a categorization of NFHS I in terms of a childcare
provider when women work aw~y from home and whether they take care of children
when they work at home. It combines both. However, it is found that around 43

percent of women combine childcare along with work (both at home and away).
Table 8.4: Primary care provider for 0-3 age group when women workprimarv data
I Type of care
DW+SHG DW+ME+SHG
DW+ML+SHG
DW+ML+ME
: provider
+SHG
! Mothers
86
84
45
56
I Grandmothers
II
34
8
29
I Fathers
2
2
3
2
I Other relatives
2
1
12
II
I Neil!hbours
0
0
5
2
. Paid maids
I
0
0
0
I Private creche
0
0
I
0
rOthers
I
I
0
I
Total
100
100
100
100

Children in the age group of 0-3 require full time care providers. Though caring for
children in a family is performed by more than one person, the above table prepared
using primary field data provides information on single major care providers. It is

found that the share of women who primarily engage in market labour is

119

comparatively less in engaging as primary childcare provider. Since the majority of

them work

away from home, substitute care is required.

It shows that in a

considerable number of cases grandmothers take up childcare responsibility when

women work outside the home. The share of husbands as the major care provider is
abysmally low.

Other childcare arrangements such as relatives and neighbours

taking care is more in the case of women engaging in market labour and micro
enterprises.
Table 8.5: Primary care provider for 3-7 age group when women work Primary d ata
DW+ML+ME
Type of care
DW+ME+SHG DW+ML+SHG
DW+SHG
+SHG
provider
81
13
82
11
Mothers
9
14
13
13
Grandmothers
5
4
4
3
Fathers
2
3
I
4
Other relatives
3
1
6
2
Neighbours
0
0
0
Paid maid
0
0
0
0
Private creche
0
0
0
0
0
Others
100
100
100
Total
100

Children in the age group of 3-1 do not require full time attendants as they go to preschool and school during day time. Yet, th~ arrangement for sending them to school
and collecting then again is needs to be made. In this age group, the share of mothers
as primary care providers is high when compared to that in the age group of 0-3. It is
noticed that the share of fathers as the major care provider increased compared to
children in the younger age group, as given in the previous table. In both the above

120

tables 8.4 and 8.5, the paid maid and private creche etc. are not the option for
childcare and this is particularly because all families are below the poverty line.
It has been found that majority of the children in the age group of 3-S go to
anganwadi centres. Being a state owned institution for pre-school children, it has
relevance particularly among the poor population and it has emerged in the focus
group discussion on childcare facilities as well. The following is a discussion on
anganwadi centres, an institution that has also undergone changes as part of the
decentratralised governance initiative of the State.
8.2. Anganwadi centres in Kerala
The ICDS (Integrated Child Development Scheme) with its network of anganwadi
centres covering more than 3000 community development blocks is perhaps the
largest women and child development programme being implemented anywhere in
the world. (Seema 2001). The ICDS programme was started in 1974 with the basis
purpose of meeting the health, educational and nutritional needs of the poor and
vulnerable infants, pre-school children, and women in their childbearing years.
Supplementary nutrition, nutrition and health education, Immunization, health checkup, referral services, and non-formal 'pre-school education are the specific
components of the program. The non-formal education of children in the age group
of3-S lays emphasis on play and creative activities aimed at the mental and physical
growth of children. The first ICDS project in Kerala was set up in 1975 in the
Vengara block in Malappuram district under the first batch of projects launched in
~e

country. Currently, more than IS,OOO anganwadi centers are functioning in the

121

State, which has a total of 152 blocks. The remaining 32 blocks have balawadi
centers set up by the state government. These balawadi centers have now been
converted to ICDS centers. It is estimated that 10.32 lakh women and children have
benefited from the scheme through 15, 60S anganwadi centers in the State in 199697 (Seema 2001).
In Kerala, the decentralized planning campaign started during 1996-1997 has made
noteworthy efforts to improve anganwadi centres, particularly the infra-structural
facilities.

In an evaluation study of ICDS in Kerala, Seema (2001), who was

involved in the campaign for democratic decentralization has noted the following:
There is a rising demand for quality pre-primary education. Owing to
high levels of literacy in the State, the demand for pre-primary
education is quite strong and widespread. This explains the
mushrooming of private nursery schools charging hefty fees and
donations for admission:' It is primarily those sections of the society
that cannot afford such expensive private schools who tum to the
anganwadi centres for their services. However, the physical condition
of the anganwadi that supposed to provide scientific pre-primary
education is very poor and unattractive. There is an urgent need to
reorient the ICDS in the State for addressing these issues that are
specific to the State and also devise means most suitable for its sociocultural milieu. Given the rich tradition of public action from below,
it is also important to think in terms of greater community control and
participation of the anganwadis. This line of thinking takes us away
from the rigid pattern of a vertical programme to a programme that is
designed in a decentralised manner:
(Seema, 2001)
Hence, there was a concerted effort in the direction of improving facilities in the
anganwadi centres. Since the study sample is drawn from the population below the
poverty line, most of the women send their children to anganwadi centres. Women in
Ithe focus group discussion have reported that anganwadi workers are overloaded

122

with many jobs being assigned to them as part of the decentralisation process, which
has resulted in significant adverse impact on their teaching inputs and spending
quality time with children. Another point that emerged out of the focus group
discussion is that anganwadies need to be more responsive to the time schedule of
women's work.

Given the fact that women engage in more than one income"

generating activity and that certain kinds of work demands working time in the early
day or late at night, the timings of anganwadi centres need to be rescheduled to
accommodate the same. Nevertheless, it has been mentioned that this cannot be seen
as an alternate childcare mechanism since it is limited to children in the pre-school
age group. The need for formal institutional arrangement for children in all age
groups has been voiced in the discussion. While discussing the possible alternate
childcare arrangement that can cater to the needs of all mothers, it is clearly
mentioned by women that the anganwadi being a top-down government programme
has little space for the same. A better alternative suggested by them is a care center
for children in all age groups run by local self-help groups charging a nominal fee.
Here, it is quite interesting to note the fact that government's role in providing such
services is no longer seen as a solution and a right of the citizens. The solutions are
community institutions that charge nominal fees and most often fall under the aegis
of self-help groups. It marks a period of shift in public and private social
responsibility .

In the context of insufficient formal

institutional arrangements, the family remains
,

the major social institution for providing childcare. As we have discussed in chapter

I five, the childcare provider and arrangements within a family varies across families
123

and changes within it as well. In certain cases it may be the mother or mother-in-law
being the care provider for a longer time period, who may stay with the daughter or
the daughter-in-law. In some other cases, the care provider and the arrangement
changes often based on the convenience of all care providers. Sometimes, the child
is sent to the grand parents and collected back daily.

No matter what the

arrangement, women have mentioned in the focus group discussion that the fact that
the primary responsibility of arranging the same lies with them is the one that causes
greater stress. Also, proper childcare centers with flexible timings are not only the
need for mothers who work outside home, such a need has been voiced by mothers
not engaged in market labour or micro-enterprises production and living in nuclear
families.
8.3. Summary
The existing institutional arrangements for childcare are assessed in the present
chapter using both secondary and primary data. It shows that the existing formal
institutional arrangements for childcare is insufficient. The anganwadi centers being
state-run institutions are limited only to children in the pre-school age group, and are
insufficient to cater to existing childcare ~equirements. In the focus group discussion,
women have expressed the need for a childcare center that accommodates the varied
work timings of mothers. Such a flexible arrangement is the needed for mothers who
are not engaged in market or micro enterprise production also.
The following chapter primaril):, discusses childcare within the family in terms of
seeing it as a caring labour and the sharing of such labour among family members.

124

When women work outside home, it is expected that the share of men in caring
labour might increase and reduce the caring responsibility of women. How caring
labour is gendered and does not work in favour of women though they earn income
is discussed in the next chapter.

125

Chapter 9

Impact of Maternal work on Paternal Care

9.1. Introduction
There has been a large chunk of literature on women's engagement in the market
labour and its impact on their familial responsibilities including childcare. While
exploring the relationship between women's involvement in the process of
production and their reproductive activities, often the underlying assumption is that
the primary responsibility of caring children lies with them. The focus, hence, is
mainly on the compatibility question on the ground that the caring responsibilities
naturally lies with the women and the analysis would be'in assessing the adverse
impact of the work participation on caring labour. The research questions are often
framed around how maternal care is compromised by their involvement in the
production process.

However, it is expected that women's participation in the

market labour would result in their greater autonomy and a renegotiation of gendered
roles within the family.

..

Hence, the attempt in the present chapter is to understand the concept of 'care' in
terms of seeing it as labour. Also, the gendered nature of this 'caring labour' is
addressed. It is expected that women's work outside home might alter the existing
division of unpaid domestic labour in favour of women to some extent. A higher
amount of paternal care is viewed as a fairer distribution of caring labour between
the sexes. Thus the pattern of paternal care and its nature and extent are discussed in

126

this chapter. A multivariate analysis has been perfonned in order to quantify the net
effect of maternal employment on paternal care. Various aspects pertaining to
paternal childcare is collected in the focus group discussions is also presented in this
chapter.
9.2. Understanding Caring labour
Studies exploring the ways in which the concept of caring labour is significantly
different from traditional economistic definitions of work is an outcome of the
feminist scholarship in the recent past (Folbre 1994; Himmeweit 1999). The concept
of caring can be described as a type of work that requires personal attention, services
that are nonnally provided on a face to face or first --flame basis often for people who
cannot clearly express their own needs, such as young children, the sick or elderly.
Care is given both freely and through the market in the sense that 'care services' are
mediated by the market and 'care giving' is formally voluntary. But in addition to
describing a type of work, caring labour describes an intrinsic motive for perfonning
that work - a sense of emotional attachment and connections to the persons being

cared for (Folbre, 1995). Hence, it is closely related to another concept that has
received more attention in sociological literature i.e. emotional labour.

Folbee

defines it as "labour undertaken. out of affection or a sense of responsibility for other

people, with no expectation of immediate pecuniary reward" (p.75).

127

Badgett and Folbre (2002) make a distinction between 'care services' a type of work

and 'caring motives', which are intrinsic to the worker. They argue that many,
though not all, care services are inspired by motives more complex other than
monetary or instrumental concerns. They involve a sense of connection with the
care recipient than may be based on affection, altruism or social norms of obligation
and respect.
Caring as a labour developed by the feminist scholars is distinctly different from the
Marxian and neoclassical approaches to the labour process.

In the Marxian

terminology, the necessary labour is that labour which is necessary to 'maintain and
reproduce' the worker and / or his labour power. Bubeck (1995) argues that caring
labour is necessary labour on the ground that our nature is such that we need livable
surroundings, food that needs preparing and clothes that needs washing, that we need
to be looked after as children, and we need others to care for us if we are sick, frail,
or disabled, caring labour is certainly part of that labour which is intended by the
notions of necessary labour even if it was not intended by Marx. Also, necessary
labour is labour that allows us to maintain and reproduce ourselves, i.e. meet our
own needs, whilst caring labour is typically and to a large extent geared towards
meeting other peoples needs. A lot of c3ring labour is labour that we could not do
ourselves if we were in need of its benefits i.e. children could not bring themselves
up, nor could somebody bedri!lden be able to care for herself. However, it is
nevertheless labour that needs to be done.

128

Also, the Marxian tradition locates caring labour within the family and community

and it is basically production for use rather than for exchange and hence outside the
purview of the market. Since, there has been no dialectical development in the
history of caring labour, it comes under the category of unalienated labour. Here, the
implication is that it is more enj.oyable or fulfilling than any other labour. However,
feminists scholars have focused their attention on the contradictory dimensions of
care as an activity that is frustrating as well as rewarding (Folbre 1994; 95)
The neo-classical tradition that takes utility maximization as a major concept can
accommodate an array of motives for labour. However, it does not conceptualise
caring labour as it takes utility function as exogenously given and there by ignore
how they are socially constructed. . It interprets caring as an exogenously given,
biologically determined altruism I.

9.3. Geoder aod cariog labour

Caring in all its aspects is highly gendered. Caring as an activity, disposition, and
attitude forms a central part of cultural conceptions of feminity. Hence, it is firmly
located in the women's side of the sexual division of labour. Feminist scholarship
emphasizes the remarkable similarity between women's responsibilities for care in
the home and their responsibilities for care in paid jobs, which is cited as the reason
for undervaluing of the paid care. Badgette and Folbre (2002) argues that social
norms which define women's

~eater

responsibility for providing both remunerated

, Although Becker has written on endogenous preferences, they are analysed in a framework that
ilflores both the social construction ofpreferenees and forms of collective action to enforce them.
S" Nancy Folbre (1993) for a detailed readil\g.
\

129

and unpaid care are not benign nor can that responsibility be explained as simple
altruism. Instead, they see it as socially imposed altruism from which men as a
group have much to gain. Hence, feminist theory emphasises the coercive dimension
of social norms of masculinity and feminity, describing norms as important elements
of gendered structures of constraint (Folbre 1994; Kabeer 1986).
9.4. Patemal care as the distribution of caring labour
One of the major components of caring labour is the care of children. In the social
division of labour, the caring of children is entirely a feminine act, which is highly
informed by the notions and institution of motherhood. Recently, in the western
industrialised countries certain emerging movements of men such as 'Taking
Parenting Public' are drawing the attention to the question of fathering.
Subsequently many research studies have focused on socio- politics of fatherhood
and how the practice of fathering is situated within a network of social structures and
institutional forces including the welfare system (Hobson 8, 2(02).
A study conducted at the National Centre on Fathers and Families, university of
Pennsylvania, developed following multi- dimensional indicators to understand
fathering.
Fathering Indicators Framework (FIF) Categories and Operational Definitions

I.Father presence
Ibis is a three-part process involving father engagement, availability, and
responsibility in relationship to children

130

1. Care giving

Providing nurturance and performing routine tasks necessary to maintain children's

emotional well-being, physical health, and appearance

3.Children'social competence

Actively engaging with children and their academic achievements and other
activities in developing and enhlmcing their social competence and academics

4. Cooperative parenting

Fathers, mothers, and other caregivers establish a supportive, cooperative and

interdependent relationship aimed at optimal child development

5.Fathers' healthy living

Providing a role model through healthy lifestyle, education, and appropriate social

behaviours that teaches work ana personal ethics, as well as social norms, to help
children grow emotionally and become productive members of society

6.Material and Financial Contributions

Engaging in consistent activities that provide material and financial support to
children

131

Better sharing of childcare between the sexes shows fairer distribution of caring
labour within the family. It is in this contest that an analysis of paternal care is
carried out in order to understand the determinants of paternal care. In our present

analysis, we pay attention only to the physical care giving component of paternal
care.

Figure 9.1: Average dally time spent on chlldcare by fathers

DW+MI.+ SHG+ME
DW+ML+SHG
DW+SHG+ME
OW+SHG

0.5

1.5

2.5

3.5

Hours

.-----------------~----------------------------------~

Figure 9.2: Average daily time spent on childcare by

grandmothers and others

DW+ML+ SHG+ME

!
i!

DW+Ml+SHG

DW+SHG+ME

DW+SHG

Hours

132

The figures given above clearly show that as the maternal workload increases, the
amount of time of paternal care decreases and the childcare provided by
grandmothers increases. This behaviour needs to be analysed and understood. The
following analysis tries to understand the extent of paternal care in the context of
maternal work.
9.5. Paternal care and maternal work - An Analysis
Information has been collected on the nature and extent of paternal care.

The

father'S care was measured on,an ordinal scale and women were asked to specify
their position in the ordered scale that varies from very high, moderate to no support,
Information on the amount of time fathers spend on childcare, the type of caring
activities they perform etc. have also been collected. An analysis has been attempted
in order to understand the factors determining paternal sharing particularly with
reference to the role of women's market labour involvement.
. 0 ft paterna care bbV materna wor k
Tbl91P
a e . : ercental!e d'Istn'b ullon
Very high
Moderate
Nil
DW+SHG
37
30
33
DW+SHG+ME
34
27
39
DW+ML+SHG
34
42
24
28
33
39
DW+ML+ SHG+ME
31
Total
31
38

Around 40 percent of women report that the share of childcare on the part of their
husbands is nil. This applies to all work typologies. Quite contrary to our
expectation, the incidence of 'very high' level of paternal care and support is less
when women combine market labour and micro enterprises activities along with
domestic work as compared to women who perform mostly domestic activities and

133

SHG activities. A quantitative analysis is carried out to understand the factors

determining such an unanticipated relationship between maternal work and paternal
care.

In the present analysis, paternal care is considered as a function of certain identified

variables as follows.
Paternal care

f (husband's income, education, maternal work status, nature of

family, age of the child, substitute care, E)

Husbands income = monthly in~me in rupees

Education = completed years of schooling

Age of the child = age in completed years

Substitute care - I if grandmother available and 0 otherwise

Since the dependent variable is an ordered choice variable, an ordered probit model
can facilitate an analysis of this data. The ordered probit model is formulated as
follows:

y.
Y.

Wx +E

is unobserved and can be thought of as the underlying tendency of an observed

phenomenon and it is assumed that E is normally distributed with zero mean.

f3' X is

a vector of the respondent's characteristics. What we do observe is

y=o,

if y. :S IJo (=0)

134

y = 1,

if

Y =2,

if Jil 5; y. 5; J12

y=J,

ify j-I 5; y.

5; y.5; Jil

where y is observed in J numbers of categories and !lS are unknown threshold

parameters separating the adjacent categories to be estimated with

13.

The first

threshold parameter JlO is typically normalised to zero so that we have j-l parameters
to estimate.

With normal distribution, we have the following probabilities:

Prob (y = j) = cP (/lJ - PX) - cP (......1 - PX)

where ( ) denotes standardised cumulative normal distribution.

In our study, the dependent variable is an ordered choice with three categories. Then
we have,

Prob (y=O)

cP (-WX)

Prob (y= 1)

cjl(JiI -

WX) -

cjI(-W X)

Prob (y=2) = 1- cP (Ill' WX)

In the probit regression, the probability of y is not a linear function of Z and it is a
cumulative normal function of Z. Hence, coefficients cannot be interpreted as in the
case of OLS. Here, we need to estimate the marginal effects which is basically the

135

effect of a unit increase in X I on the probability that y= I, when all other independent
variables are held constant at their means.

..

T a ble 92 E 5 ti rna t eso ftb e ord eredI pro b'It rnod e I - extent 0 ( aterna care
Coefficients
Mean of X
Variables
P values
Constant
0.841
0.008
-0.012
0.005
Husband's income
1650
Age of child
-0.333
0.000
3.8
0.162
0.468
8.4
Husband's
education
-0.463
Maternal work
O.oJI
0.312
-0.341
0.012
0.425
Substitute care
0.96
0.000
III
N=421

chi-square (df 5) = 8.2 III is the threshold parameter

. bles on extent 0 f: paterna care

T a ble 93
. I e(Iiect 0 (.ID depen d ent varIa
. : M arl!lDa
Prob (y=1)
Prob (y=2)
Prob (y=O)
Variable
+0:002
-0.004
-0.004
Husband's income
-0.007
+0.003
+0.009
Age of the child
0.016
0.382
Maternal work
0.024
0
I
Change
Substitute care
0
I
Change

0.051
+0.035
0.025
0.070
+0.045

0.387
+0.005
0.0315
0.0219
0.095

0.006
-0.0017
0.320
0.186
0.134

9.S.1. Discussion

Husband's income and age of the child are continuous variables, while others are
dummy variables. Coefficients shown in the Table are all significant at 5 percent
level except husband's education. The positive and highly significant

III

shows that

the dependent variable is indeed ordered.

The results indicate that when there is a unit increase in husbands income, the
probability of their not caring children increases by 0 .002 and that of highly caring

136

decreases by 0.004. Hence as their income increases, the fathers are less likely to
involve themselves in childcare activities. As the age of the child increases by one
year, the probability of not caring by the father by the father decreases by 0.007 and
that of being cared increases by 0.009. With regard to maternal work, when mothers
work outside home for cash in~ome, the chance of being cared by fathers decreases
by 0.017 and that of being not cared increases by 0.035. When there is substitute care
available at home, the probability of father's care decreases by 0.134 and that of not
caring increases by 0.045

The results show that when children are small and their mothers go out for work,
they are less likely to be cared for by fathers and are more likely to be cared for by
substitute care providers at home. Since the burden of child caring labour is more in
the early ages, it is clear that there is no fairer distribution of the same at the familial
between the couples. The same has been discussed in the qualitative round and
described below.

9.6. Qualitative data from focus group discussion

The child caring work of husbands is perceived by women as a help, which they seek
only when it is 'required'. In most of tfle cases when the question on sharing of
childcare was asked, the response was that the help is basically a need based. If
mothers are not able to perform the same, if nobody at home to share the caring
activities, the 'need arises'.
components of childcare.

'It cannot be seen as a regular sharing of certain

Even for women who come under the category 'very

high', claiminl1 hieher del!ree of care sharinl1 comes under the 'need based' support

137

seeking behaviour. Hence, when women go for work. outside the home, a substitute
care provider takes care of the child. In most of the cases, they are mothers or
mothers in law of the women. It is arranged in varied ways, which are convenient for
both the parties. Mothers may temporarily shift to daughter's place to take care of the
child or may be sending the child to the natal family in the morning and collecting
back in the evening or it could' be the child staying with natal family for a fixed
duration of time etc.
It is also emerged in the discussion that fathering is more of an overall supervisory
role in the childcare regime. Fathers supervise whether children get proper food; they
wear neat dress, whether medicine is given on time when they are sick, whether
children are studying well and getting good marks in the examination etc. This
supervisory role of childcare by fathers falls under the category of passive care.
When women do the supervision of the entire child caring labour, it is more
informed by their responsibility as primary care providers. Whereas in the case of
fathers, it arises out of their authority to monitor the caring work as head of the
household. In the case of an identified fault, mothers are only often being accused of.
Women also bear the feeling of culpable and totally justifiable to be accused of when
there is a major slipup takes place. Many women have reported in a positive manner
that their husbands are too strict on such things cannot tolerate any slip up in
childcare. Fathers take the final decision regarding disciplining and punishing
children. It is also mentioned by many women that since they start earning income,
they do purchasing dress and other needful items for children without consulting to
Itusbands each time.

138

Yet another aspect is the quality time of fathering. It is observed from the discussion
that while fathers spent time with children, mostly they engage in conversations,
playing, story telling etc. This might be having significant impact on the
socialization of children and helps them learn new ideas and many new things
around them. The impact of such socialization on children is a researchable issue that
the future research can be addressed.
9.7. Summary

It is expected that women's work outside home might result in a greater degree of

sharing of caring labour with other male members of the household.

It is also

attempted to understand the caring labour conceptually and its gendered dimensions.
The analysis made has shown that the degree of sharing of childcare responsibilities
by men are very limited. Women go for work outside the home, in a significant
number of cases, either her mother or mother in law or any other female member of
the household takes up the responsibility of childcare. Children above the age of 3
years are sent to pre-school informal learning centers, anganwadies. This result is
also in corroboration with the low levels of female autonomy measured in the private
sphere. It indicates that women's work outside the horne is not realigning the labour
division within the domestic chores by involving male members considerably.

139

Chapter 10

Working life and family dynamics -Reflection on Case

Studies
10.1. Introduction
In order to further the understanding on the dynamics of intersection of work, gender
and family under study, case studies have been conducted. (Cases are appended as
annexure II). Ten cases have' been selected while conducting the questionnaire
survey. Though certain responses were of considerable relevance to the central
theme of the study, the questionnaire format does not accommodate a deep probing
that enable us to understand the dynamics of the research issue under investigation.
Such occasions were noted and ten cases were selected among them for detailed
probing. The focus has been given on the aspects such as work women do outside
home, their views on kudumbashree, compatibility between domestic work and work
outside home, understanding on autonomy and decision-making, childcare
responsibilities and institutional. arrangements. The following is a reflection on the
selected case studies, which is done against the backdrop of the analysis that has
already been carried out using both quantitative and qualitative data from primary
and secondary sources and the results that have emerged out of them.
10.2. Cbanging dynamics of work
The income women receive from self-help micro enterprises is considerably low
when compared to existing other wage labour. An early impact study of the CDS
experiment in Alleppey Munic!J?ality upon which the Kudumbashree was modeled

140

pointed out that out of the sample of 1279 women surveyed, nearly eighty per cent of
the members do not think that the income generated has been adequate. Only 3.6 per
cent reported strongly about the income benefit through income-generating activities.
(Oommen, 1999). The situation seems to continue at least in many areas. A recent
study points out that in its study area, the earnings from employment created through
SHGs are very small, ranging from Rs 500 to Rs 1000 per month and the problem of
women's double/triple burdens of work seemed little addressed (Eapen and Thomas
2005). It is found in our case study too that women who combine both market labour
and micro enterprises production are spending 14-16 hrs a day for these works.

Also, in all the cases we have studied, women working with micro enterprises have
expressed their concern over low levels of income they earn from micro enterprises
production unit.
Nevertheless, the kudumbashree micro enterprise production is viewed as a 'better
status' work in comparison to other traditional works that are often more paid.
Though many women expressed their overt concern over less income accrued from
micro-enterprises and long duration of working time, they do not want to fall back to
the wage labour. Instead, in certain cases women tend to combine both micro
enterprises production and wage labour as an initial strategy to manage income with
a future plan to leave wage labour when the micro enterprise they work on flourishes
and accrue enough income. Wage work is typically been viewed as a 'low status'
work and hence the ultimate aim is to gradually move out of the same.

, Availing loans through self-help groups is a newly opened up avenue not only for
initiating income generation activities, but also for acquiring household consumer
141

goods. Such consumerist culture and aspiration for upwards social and cultural
mobility are sky-high not among women in the study setting. Majority of women
have proudly narrated the assets and consumer goods they acquired through self-help
activities, which would not have been possible, if they had not been joined with
kudumbashree self-help groups.

Indeed, the asset building effort that women pursue and the instances of negotiating
with husbands and others in the family for the same is of special relevance. Its result
could be that the woman gets tied to what Hannah Papanek has called "family status
production,,22. This is particuiarly relevant for Kerala where social mobility
aspirations are sky-high and consumerist aspirations abound among all sections of
the people, and not just the propertied middle class.

Yet another feature of the work in micro enterprises is that it has flexible timings and
accommodates domestic labour demands of women 2J Besides, the travel cost is nil

in comparison to work that they take up in the near by city as helpers or sales girls.
Further more, since women from self-help groups are actively involved in many
development activities carried out under the auspices of gram panchayat, they have a
social acceptability as community development workers. Certainly they do have
By which she means the totality of labour the elements of which include support for the paid work
of other family members, the future paid work and status ambitions of children, ensuring status
through social activities, and religious activity. Hannah Papanek (1979). This applies quite well to the
situation of upper class highly educated Malayalee women today, and is spreading to ever-greater
number of groups including poor.

22

Mridul Eapen and Soya Thomas note that women in SHGs in the panchayats they studied, despite
the very many serious problems the researchers found, were upbeat about the SHGs:"Given the fact
that 70-75 per cent of women in Kerala are engaged in household duties, many of the SHG activities
able the poorer among them to combine roles and earn some income. Women perceive them [work
nearer to home and not demanding full time commitments) as a boon and the extra work appears to be
no burden to them." M. Eapen and Soya Thomas (2005)
23

142

more access to local political society and bureaucracy.

Precisely, realizing the

aspirations for women social' acceptance, efforts were made by Kudumbashree

bureaucracy to get acceptance for the sort of community work such as scavenging
and waste disposal under the new name 'Clean Kerala business'.
The main objective of the Kudumbashree programme is to relieve the poor women
from the clutches of the moneylenders, who charge exorbitant rates of interest. But
even after joining the programme, the borrowing from the money lenders is
continues to be high. The reason cited for this high rate of dependency over
moneylenders is the inability !o sustain their small enterprises with the initially
borroWed amount in the case of those who have initiated micro enterprises. Another
common reason to all is to meet the emergency consumption needs. It ranges from
giving gifts to marriages of relatives or neighboUrs to health care needs of the family
members. Also there are local pawns where any movable assets including gold can
be pledged. Women selected for the case studies have mentioned that they do such
mortgaging often without knowing their husbands or other family members. Further,
there are local unregistered chit funds that women are members in addition to their
weekly savings in the Kudumbashree SHGs. It indicates that though the newly
placed government institutional strueture is not replacing the traditional money
lending and managing mechanisms. Women see SHGs as a structure of the state and
they need to participate in the same for many benefits that is given by the State to the

poor.

143

Another aspect of kudumbashree that women in the selected case studies have
mention is that of the collectivity of women in the SHGs. Three women have
mentioned that it is a collective learning process for them. It gave them necessary
skills including the communication skill and negotiation abilities, primarily outside
home. Also, they indicated that the relationship among member women has
improved and it is a source of strength to them. The process of developing trust and
friendship among member women has been described by women in the selected
cases. Initially, the contlict and tension intensified among members mainly due to
the reason that such interactions and collective action were absent in their lives.
Later, they learned to be in groups and started discussing the personal matters too
and the relationships got balanced gradually. Now, it has been considered as a space
for an outlet to release their tension and to ease. Yet another benefit that women see
in kudumbashree SHGs is that it enhanced their mobility. In the narne of SHG
meetings and other related works, they can go to other places and meet their friends
and attend religious functions etc.
Instances of women identifying various social axes of power have emerged out of the
case studies analysed. A president of ADS, who received left party membership, is
aspiring to contest in the next. panchayat elections. She sees it as a major power
center in the local political scenario and a means to achieve social respectability. It
takes place in a context where more party based political interference in panchayat as
well as kudumbashree functioning takes place. Also another woman selected for case
study belongs to dalit community has mentioned that she experiences caste
domination in her group and opined that the same caste grouping in SHGs is better to

144

overcome this. It is also emerged that in the groups, those who are relatively
economically better offtake leadership roles and others get marginalized. Often, they
become the presidents of ADS and CDS.

Hence, women in the SHG groups

'.

experience various axes of social power such as caste, class and party politics, and
the dynamics of these social powers and its impact on women's work and social
position is significant.

10.3. Intensity and mUltiplicity oework and childcare arrangement

A commonly shared experience of all the cases selected is that of multiplicity and
intensity of work. The combination of different paid works such as combining micro
enterprises production and market labour, engaging in more than one market labour,
micro enterprises production and petty trade is found in the selected case studies.
The net result of all these are excruciating stress and strain for women and a
significant reduction in rest and leisure time. In toto, it more responsibility financial
responsibility to shoulder in addition to the traditional caring labour and domestic
work.
Despite the results of quantitative analysis indicating that maternal work has no
adverse consequence on child healtl) indicator, which hardly indicates that the
maternal work - childcare is a tension
free zone, women do experience the hardship
,.
of arranging substitute care. Women make all probable arrangements for childcare
that includes entrusting the task of childcare to a female member of husband's
family, most often the mother-in-law, sending the child to her natal family and
collect back on a daily basis, make her mother stay with the child in her house

145

temporality, seeking the help of neighbours etc. The kind of tension and energy
investment in making various childcare arrangements for time-ta-time became
evident from the analysis of selected case studies. When these options are not
available, they take employm~nt breaks. It is also emerged that there is more
community life in the colonies and the childcare is being shared by neighbours. It
happens so, because there are many relatives live in the neighbourhood of the same
colony.
Women explained the stress they bear in order to make childcare arrangements.
Irrespective of the work status and combination of works, majority of the women
considered for case study have expressed their need for a better childcare
institutional facility. It should take care of the flexible timings of mother's work.
Also the cost of such care ammgement has to be less and affordable for poorer
sections.

146

The idea of quality upbringing includes providing nutritional food and proper
medical facilities when children are sick. Sending children to English medium
schools is another aspect of high quality upbringing. It is found that there is no
discriminatory attitude based on gender in their expectations to provide such
education. It is being viewed. as an indicator of better social status. But the
expectation on the outcomes of such an education is gender differentiated. To boys,
it is expected that the better quality English medium education will bring better jobs
and social status. Whereas for girls, it improves their relative position in the marriage
market though the dowry might not come down always.

10.4. Autonomy and decision-making

As found in the focus group discussion, in the case studies made as well, women
narrated their experience that

~eir

enhanced decision-making power in recent times

is mainly due to their changed economic status. It may not always be an indicator of
autonomy and enhanced status but reveals the increased responsibility transferred to
women by the male income earners of the household. The share of male income to
the family maintenance has reduced in certain cases.

In many cases women

expressed the differences in priorities for expending their income. Women. Many
instances of women negotiate with men to buy land and build house, buy gold, other
household assets etc. have been narrated.
Another important aspect that emerged out of case studies is that in spite of women's
work in the self-help groups and attending the training programmes on

147

empowerment conducted by Kudumbashree, incidence of domestic violence is high.

It is more among families where alcohol consumption among men is high.

It is observed that fertility decision-making process is also slowly changing. The

quality upbringing of children as well as increased work demands push them towards
reducing fertility to the lowest level and in two case, women decided to have only
one child. This is an area that future research can focus.

148

Chapter 11

Conclusion and Policy Recommendations

11.1 Summary aud conclusion

The central theme of the present study is to understand the relationship between
women's work and child health and care, which is placed in the broader context of
the intersection of gender, work and family. The study employed an existing causal
model to examine the net effect of women's work on child health and care using both
primary and secondary sets of data. Nevertheless, the study foregrounds a broader
terrain, which specifically sees women's work in terms of its intrinsic value and
instrumental value. The intrinsic value of work sees the right to work as a right in its
own way that provides women an agency to exercise their autonomy, which
transforms them and influences family members and the community.

The

instrumental value of work measures the anticipated utility of work in terms of the
changes that it would make in other variables of interest such as fertility and child
mortality. Hence, it received wide attention in the area of demographic research and
public policy advocacy. The role of ,women's education and work participation in
reducing fertility and child mortality has been widely acknowledged and micro
studies assessing the influence of female work status in the context of developing
countries has been well documented. Many policy prescriptions are also put forth

based on the strength of instrumentality of women's work. Studies examining the

impact of women's work on child health and care, however, are less conclusive

149

about the direction and net effect of causality. It results in giving more weight to the
social context in which women work and the nature of support mechanisms available
for childcare.

Also, the third vantage point through which the study has been

pursued in understanding caring labour and to assess whether women's work outside
home has significantly altered the gendered alignment of caring work within the
domestic domain.

The methodology of the present study is interdisciplinary in its approach. Though the
central theme of the present study, the relationship between women's work and child
health and care is articulated within the logical framework of demographic research, it

borrows from feminist scholarship particularly while discussing female autonomy and
caring labour.
While acknowledging the value of quantitative data and methods in documenting
certain inequalities, the methodological stance adopted in the present study
recognizes the importance of gender as a social construct that goes beyond an
understanding of inequalities in individual level variables and its all pervasiveness as
an organizing principle of all societies that could be better captured using qualitative
data.

Hence, the complementarity. of qualitative and quantitative methods is

acknowledged and adopted. This is not only in the case of understanding gender but
also holds good in the case of understanding various social aspects of women's life
such as work and childcare. The specific trajectory of inquiry adopted in the present
study is to quantify the formulated relationship and then to seek qualitative
information to either supplement or to reconceptualise the relationship under study.

ISO

The sources of data employed in the study are both secondary and primary. The
National Family Health Survey (NFHS 1&11) data are used to examine the impact of
women's work on child health outcome, the morbidity incidence, using multivariate
analysis. Also, the second round of NFHS data provides a measure of female
autonomy and it has been used to understand the factors that determine autonomy as
measured in the NHHS II. The primary data include both quantitative questionnaire
survey and qualitative information collected through focus group discussions and
case studies. The field setting for the primary survey is selected from
Thiruvananthapuram district of Kerala State. Two gram panchayats with female
work participation rates closer to state average viz. Pallichal and Vellanad are
selected.

It is recognised that the nature of work women perform and the social context in
which their working lives are situated would be significant in understanding the
intersection of gender, work and family in any given context. Hence, the nature of
women's work is explored in the study setting.

It is found that there is a considerable 'amount of economic restructuring of working

lives of women around micro-credit and micro-finance activities, using self-help
groups as the institutional mechanism. The State initiative of Kudumbashree. the
poverty eradication mission, organised poor women all over the State into self-help
groups called Community Development Society (CDS), a three-tier structure
federated at the village pachayat level and brought them under the local governing

151

structures. This linking up of 'women SHGs with local governments was a part of
participatory planning process carried out in the State as part of the decentralisation
process that includes devolution of powers, responsibilities and finances. It is being
projected as an effort towards gendering the planning process by earmarking 10
percentage of total plan funds to women as a separate an component, named as
Women Component Plan (WCP) and to form self-help groups under the local
governments to strengthen participatory planning as well as poverty reduction. It is
also expected that both these steps would enhance the empowerment of poor women.
(Issac et ai, 200 I).
Therefore, SHGs have become legitimate institutions at the local level to receive
funds from various sources, such as all anti-poverty programmes of the Central and
State governments, funds from the Women Component Plan, mobilizing finances
through other sources, including commercial banks etc. for planning and carrying out
income enhancement projects for poor women. As a result, self-help activities have
become an integral part of poor women's lives. They engage in Kudumbashree in
varying degrees such as engaging in self-help activities such as thrift, credit,
attending meetings, attending gram sabhas (village assembly) and other meetings
being organized from time to time. 111 addition to self-help group activities, some
women engage in micro enterprises production as a group based activity. The result
of this pattern of restructuring has been analysed in detail with regard to their
income, autonomy and domestic responsibilities including childcare using both
primary quantitative and qualitative data.

152

It is found in the time use survey of women that they take up multiple activities in

addition to their domestic work. Hence, a categorisation of work that accommodates

these multiple tasks have been developed. Though the intensity of work has not been
adequately measured, women in the focus group discussion have expressed the same.
The following work typology has been adopted based on the actual time use.
I. Performing domestic work and self-help group activities
2. Performing domestic work, self-help activities and engaged in micro
enterprise activities
3. Performing domestic work, self-help group activities and market labour
4. Performing domestic work, self-help activities, market labour, and engaged
in micro- enterprise activities

The time allocation for work varies considerably based on the combination of types
of work women carry out. It is also found that a reduction in the time spent on
childcare occurs as the number of activities increases on the work front.
Nevertheless, the quantitative analysis carried out has found that this reduced amount
of time that mothers spend on childcare does not significantly affect child health in
terms of morbidity incidence.
The analysis of secondary data reveals that when compared to individual level
variables, such as female education and work status, the indicators of socioeconomic context such as sanitation and drinking water facilities do have a
significant bearing on child health.

153

The analysis of primary data found that children of mothers involved in the market
labour in addition to the dome.stic work have more chance of getting affected by
diseases compared to the children of mothers belonging to other categories of work.
Children of those who belong to the non-scheduled castes and tribe community and
living outside housing colonies do have significantly less chances of getting affected
by illness. The study sample includes the SC/ST population who have less
educational achievements, are mostly involved in market labour activities such as
bamboo and reed work, and live in housing colonies. This result has revealed the fact
that the category of poor is not homogeneous and there are sections such as poorest
of the poor, further marginalise? in terms of their caste and other social background,
who face more deprivation than others. The child health status of this section is
significantly affected by their deprived socio-economic status.
On the whole, these results reveal that the socio-economic context within which

children live is more important in determining child health and care in comparison to
individual level variables such as maternal education and work status. This result has
to be viewed in the light of policy prescription towards enhancing girls' schooling
and female employment for improving child health in the general context of
globalisation. It has been argutd thai educated mothers are better able to access
medical services and follow advice, they have greater power within the household to
ensure that sick children receive care and engage in childcare practices that enhance
child health. Hence, there is an increased importance of girls schooling in the realm
of policy formulations. In fact, this better educational status of women always acts as
a proxy of the socio-economic context in which they live, especially in the setting of

154

developing countries. Advocating girl's schooling for better child health has certain
inherent problems. On the one hand, it considers educating girls - future mothers is seen as an important pre-requisite to enhance child health and thereby shifts the
burden of dealing with childcare (including dealing with diseases) from the society
as a whole to women in the poor households.

The problem here is that the

investments in girls' education is not being articulated on equality grounds and as a

matter of right.

This may be viewed in the backdrop of globalisation and the

declining trend in investing on public health. Economic liberalisation with structural

adjustment policies that advocate reductions in public investment including in the
area of sanitation and drinking water would have serious adverse consequences on
child health outcomes, and it diverts the attention from the need to invest in public
health care facilities such as sanitation and water supply. Thereby, it caters to the
ideology of neo-liberal institutions by focusing on the role of mothers in enhancing
child health and shifting the boundaries between public and private responsibilities
for care of children.
Recently, there has been a policy shift in the State of Kerala in women's employment
generation towards self-help group activities. Self-help group activities are more of
self-employed work and there is no sOcial security attached to the work. In addition,
there are family health insurance packages providing for the poor women of SHG
groups that takes care of the health cost at the household level. However, this again
needs to be viewed as another method of shifting the boundaries of social
responsibilities from the State and the men in families to poor women. The strong
message that Kudumbashree gives to women and society in general is that of

ISS

'prosperity in the family through women'. This gives a newly assumed responsibility
for women, which is carefully put forth as an extension of traditional caring labour,
i.e. providing economic prosperity through self-help income generation.
It is attempted in the present study to examine the impact of female autonomy on

child health and care. The causal model that explains the relationship between
women's work and child health and care uses female autonomy as an intermediate
variable in the association. Women's work may translate into greater control by them
over household resources and helps them to access better health care and nutritive
food for children. Also, the greater exposure they get from the outer world might
help them in accessing information for appropriate and quality childcare. All these
might work positively in impro,\ing child health. However, analysis ofNFHS II data
has shown that female autonomy adversely affects child health. Hence, an analysis
of factors that determine female autonomy is carried out in order to understand the
characteristics of women's autonomy that are measured in NFHS II. The analysis has
shown that women belonging to backward communities with low living standards
and working have more autonomy. The positive association found between female
autonomy and child morbidity is mainly due to this reason. This result is in
corroboration with the narratives collected on female autonomy in the qualitative
rounds.

Many women have ?,pined that their enhanced decision-making power

within the domestic domain is primarily an outcome of the 'neglect' by their

husbands and it is an outcome of their new income earning status, however small the
income may be.

156

Hence, a critical reflection is made on the causal framework that suggests female
employment enhances their autonomy, which in tum promotes child health and care.
The conceptual understanding on female autonomy and empowerment is also
reviewed in this context.

The NFHS II has measured autonomy in terms of

household decision-making and mobility indicators that hardly reflect the agency
that women exercise from the feminist viewpoint, an alternate measure of autonomy
is constructed using primary field data that measure the female autonomy in both the
private and public spheres. The relevant indicators have been identified by women
members of SHGs and it has been finalized in consultation with identified key
informants. It has been found that the measured autonomy in the public sphere is
higher in relation to that in the private domain. It is also noticed in the analysis of
factors determining female autonomy in the public sphere, the women who do not
belong to Se/ST with extended family structure, and engaged in both market labour
and micro enterprise production, have more autonomy in the public domain. It
reveals that higher caste status and support from the family help them in participating
in activities at the local level. Whereas in the private domain, most of the identified
explanatory variables are found to be insignificant except the family structure. The
nuclear family structure gives more autonomy to women in the domestic domain.
This result is quite interesting and needs further reflection. It shows that the elder
members of the extended family support women's participation in the kudumbashree
and panchayat activities, which'is a legitimized social space whereas their presence
in the patriarchal family structure excludes young women from key decision-making
within the family. Here the question of individual vs. collective female agency is

157

relevant. The agency that women assume from their participation in the public does
not empower them to negotiate with the patriarchal norms and structures within the
family. This might be due to the reason that the public domain they enter is a Statesponsored and hence highly legitimized space, and existing gender norms and values
are hardly challenged there.

'.

Women traditionally do all unpaid caring labour including childcare. Hence, a

conceptual understanding on caring labour and its gendered nature is made. It is
expected that women's work outside the home significantly alters the existing
household division of labour on gendered lines in favour of women. An analysis has
been carried out to understand the sharing of caring labour with male members of the
household and it is found that the share of male involvement in childcare is minimal.
When women go for work, other female members, mostly grandmothers, take care of
the child. Women perceive the ~hild caring work of husbands as a help, which they
seek only when it is 'required'. In most of the caseS when a question on sharing of
childcare was asked, the response was that the help is basically need based.

If

mothers are not able to perform the same, if nobody is at home to share the caring
activities, the 'need arises'.

It cannot be seen as a regular sharing of certain

components of childcare. Women have also mentioned that their work in the micro
enterprise and self-help activities is approved of by men on the ground that child care
and household work is not compromised. Managing all these activities causes more
stress and strain to them.

158

An evaluation of the existing institutional arrangements for childcare is made. The

family remains the major social institution for childcare and the role of other
institutional structures is limited for children below the age of three. The presence of
a grandmother as a substitute care provider is found to be significant in the analysis.
For pre-school children above !he age of three, anganwadi centers are widely used.
However, in the focus group discussion women have mentioned that SHG womenrun childcare centers need to be opened and it should have flexible timings to
accommodate the long working hours of women. Also, it is found that childcare is
shared by other families in the housing colonies of SC/ST communities. This
indicates the significance of social networks in providing childcare.
To conclude, the present study illustrates that though there is no statistically
significant relationship between women's work and child health, either at the macro
or micro level, women's working lives are substantially constrained with child caring
responsibilities in varying degrees. Also, it shows the significance of the social
context, more specifically the infrastructural facilities in determining child health
outcome. Female autonomy is found to be insignificant in determining child health
outcomes. Critical reflection on female autonomy reveals that the understanding on
autonomy and its measure employed in the demographic studies is highly
constrained by gendered ideas that are significantly different from the feminist
notion of female agency that women acquire, and the independent decision-making
by women to resist and negotiate with patriarchal forms of power.

Finally, the

reconstitution of poor women's working life around micro finance and micro
enterprises is substantially refiguring women's work, their autonomy and childcare
159

arrangements. The ways in which gender figures and refigures in this process is
important in understanding the nature and dynamics of this process. Hence, it is
argued here that using multiple data sets of quantitative and qualitative information
and incorporating gender as the analytical category that cuts across individual
relationships and social institutions provides better insight into the research problem
under investigation. Therefore, it is suggested that the future research in this area
would benefit if use is made of this approach.
11.2. Future Direction for Public Policy
In the light of the findings of the present study, certain public policy and advocacy
directives are suggested that could be pursued at different levels.
The study shows that though maternal employment does not result in child morbidity
outcomes, it certainly constrains women in finding alternate care arrangements as
well as compelling them to pj:rform most of the work in the childcare regime.
Alternate care arrangements mostly stay within the family relationships, which are
often seen as substitute care and help to mothers and not as the responsibility of the
family at large. Also, mothers who are not engaged in market labour and micro
enterprise production state their need to have public childcare institutions as and
when required.
Hence, an alternate public institutional care arrangement need to be devised.
Anganwadi centers are limited to children only in the pre-school age group.
Arrangements for children in the other age groups also needs to be addressed. As we
have seen, when women start taking up more and more income generation activities,

160

the duration of their working time gets longer and the childcare centers needs to be
open during early morning and late evening hours also. Such flexible timings would

be beneficial to all mothers regardless of their work status. Since panchayats do

have funds for development projects for the aged as well as for children, the old age
homes and childcare centers could be designed innovatively as a single unit that
enhances the interaction between elderly people and children. This would be
beneficial to both the groups anll different interactive sessions such as story telling,
interactive plays etc. can be developed locally.

Since it is reported by women that the income they generate is not sufficient enough,
the state government should review the micro-finance activities of Kudumbashree.
Hence, the project viability in terms of its economic benefit to the members of the
group has to be evaluated by Kudumbashree before initiating the same. In addition,
the increased working time along with the burden of domestic responsibilities are
places women in a more constrained situation. It is also found that the sharing of
caring and other domestic respOnsibilities by male partners is limited. Furthermore,
women have to shoulder more financial responsibilities as well as a result of men
becoming passive in finding employment in the rigid labour market structure that
Kerala exhibits in the less educated

~age

employment seeking labour force. The

training component of Kudumbashree needs to make efforts in understanding the

dynamics of these gender issues in the context of restructuring of women's economic
life, and appropriate strategies need to be devised.

161

Women's movement and the NGOs working for the betterment of women in Kerala
need to take serious note of these issues and must realize the critical space that they
have to keep with the State in terms of working as a group, constantly reflecting and
negotiating with the State on policy issues pertaining to women, economic work,
household labour and caring labour. More importantly, they must assume the role of
an active constituency in sensitising women on gender issues and making them
constantly reflect on it, in order to challenge and negotiate with patriarchal power
both in the private and public domains. Strategies for promoting paternal care and
sharing of other domestic labour between men and women have to be developed and
pushed fOlWard.
It is observed that the existing public - private divide in care and health
responsibilities is slowly waning and getting redefined. Self-help groups are taking
up health insurance and the responsibility of family health care is slowly falling on
the shoulders of women. Moreover, women have reported that their responsibility of
family maintenance is also enhanced as a result of their engagement in
Kudumbashree self-help groups.

Thus, the promotion of self-help groups by

Kudumbashree and the placing of women at its heart, as the individual solely
responsible for family welfare

ne~s

to be critically examined, and alternate

strategies to engender the development process have to be worked out and negotiated
with the State by women's movements and NGOs working towards bridging gender
inequalities.

162

Annexure I
Profile of the study setting
The Vellanad and Pallichal panchayats selected as the study area are having
population of 40375 and 32751 respectively (Census of India, 2001). The male
literacy rate is 81.2 and 81.62 in Vellanad and Pallichal respectively. The female
literacy rates in these two panchayats are 76% and 76.65%. Both the panchayats
show a similar level of infrastructural facilities in terms of schools, hospitals, banks,
roads etc. The share of SC/ST population is more in these selected panchayats
compared to other panchayats in the district viz. 21 % and 17.6% respectively
(Development Report, 1997).
The decentralization programme initiated in the state of Kerala since 1996 has
opened up space for women in the public sphere. 10% of the plan funds is set apart
for development projects for women and promoted active participation ofthem in the
planning and formulation of the same. Also constituted Jagrata samities under the
panchayat to address the issue of violence against women. The initiation and
promotion of self-help groups

~nder the local governments was an effort to enhance

women's economic participation. It includes both thrift and credit activities of the
members and micro enterprises for income generation activities. Since our study
sample is selected from the Below Poverty Line (BPL) list of the panchayat, all the
respondents are members of the self help groups. Some of the members are involved
only in thrift and credit activities and other development programmes organized by
the group. Another section set up micro enterprises and carrying out economic
activities.
163

In the study sample, the share of SC/ST communities is 34 in Vellanad and 28.7 in
Pallichal. The average years of education is 6.13 in Vellanad and 7.26 in Pallichal.
Women in the study sample take up variety of economic activities such as rubber
tapping, bamboo and reed work, sewing, made servant etc. The average monthly
income of respondents from Vellanad is Rs.1521 and in Pallichal Rs.IS6S.

The demographic profile and other background characteristics of the women selected
for the study is provided below.

A~ge C omposltJoo 0 fW Ol)leo

All
Age category
<19
20-24
25-29
30-34
35-39
40+
Total

Vellanad

OJ
IS.7
35.S
2S.4
13.7
3.0
100

Pallichal
0.6
IS.1
32.9
30.4
15.2
2.S
100

The mean age of women in the sample selected us 29.12 years and the mean of their
husbands is 34.07 years. The table AI.I shows that around 55% of the respondents
are below the age of 30 years in both the panchayats. In the age composition, there
is not much difference between s.elected panchayats.

AI.2. Caste Composition of the respoo d eots

Pallichal
Vellanad
Caste
2S.7
34
SC&ST
25
27.4
OBC
46.3
3S.6
Forward Caste
100
100
Total
The table AI.2 indicates that the share of SC&ST communities are more in Vellanad

in comparison to Pallichal.

164

A1.3 Na tu reo fH OUSID2

Categories

Plinth Area

<300
300-500
500-800
BOO-1000
Above 1000

Total
Concrete
Type of roof Tiled
Thatched
Others
Total
Cement
Nature of
Plastered with cow dung
flooring
Others
TQtal
Ownership
Own
Rented
Others
Total
Toilet in the house premise
Sanitary
Within house compound
facilities
Open defecation
Total
Tap in the house
Well in the house
Source of
Tap & Well in the house
drinking water Tap in the public stand pipe
Distant well
Total
Electricity
Source of
Kerosene
lighting
Total
Cooking gas
Source of Fuel Firewood
Kerosene
Total

Vellanad

Pallichal

20.9
41.2
23.3
13.2
1.4
100
17.7
39.3
34.2
B.7
100
26.1
70.8
3.1
100
83.6
3.7
12.B
100
78
4.4
17.6
100
I

4
60
36
0
0
100
36
26
32
6
100
70
4
26
100
BO.2
4.4
15.4
100
24
72
4
100
8
34
4
36
18
100
90
10
100
20
62
16
100

72
.7
13.5
12.B
100
67.7
32.3
100
12
60
78
100

Often the criticism against the income criteria to assess the economic status of
individuals is that it hardly reveals the real picture. As a result, non-income criteria
in tenns of assets, basic needs such as drinking water and sanitation facilities etc. are

165

being widely employed to understand the levels of poverty. The above table shows
the non-income criteria indicators to measure economic status. From the table, it is
obvious that all the indicators are showing low economic status of the selected
households and hence they belong to BPL families. It is noticeable that in the case
basic facilities also both the selected panchayats are similar. Hence the following
tables provide combined data of both the panchayats.
A1.4. Al!e comoosition of children
Age group
Children in the alZe I!rOUD 0-3
Children in the ae:e grOUD 4-7
Total

FreQuencv Percentage
247
42
343
58
590
100

The table above shows that 58% Qf children of our sample women belongs to 4-7 age
group and 42% belongs to 0-3 age category.

AI.S. Educational backl!round of the resoondents

Levels of Education
Percentage distribution
Women
0.7
Literate with no formal training
5.3
PrimarY
Uoner Primary
15.0
15.7
Hie:h School
lOth Pass
44.7
13.3
Plus two
0.3
Diolomalcertificate course
4.0
Dee:ree
0.7
Post graduationIProfessional COfH"Se
100
Total

Percentage distribution
Men
1.7
8.6
21.9
18.2
41.1
6.2
0
2.1
0.3
100

Table AI.5 indicates that around 60% of women are having 10 - 12 years of
schooling. Kerala is well known for its high levels of achievements in social sectors
including education. The attainment in levels of education is more among women
compared to that of men across all levels. Researchers (MukheIjee and Isaac 1994)

166

have pointed it out as one of the major reasons for high rates of female
unemployment in Kerala. Attainment in education has also lead women to look for
work having high social status oompared to agricultural labourers and wage workers.
As a result, the waiting period for ideal job tend to increase considerably. In urban
and semi-urban areas the chance of getting low paid jobs in service sector is more
whereas in the rural areas, the only avenue for employment is traditional sector
employment. The kudumbashree, state poverty eradication mission has provided
ample opportunities for setting up micro-enterprises for women in the state of Kerala
and its dynamics and impact are discussed in detail in chapter Ill.

..

AI6LaD dOw Ders h'liP

Ownership category
No land
Below 5 cents
5-10 cents
10-20 cents
20-50 cents
Above 50 cents
Total

'. Percentage distribution

5.3
21.5
31.5
29.2
11.7
3.7
100

It can be seen from table AI.6 that around 85% are having below 20 cents of land
and it also shows the low-income status of women selected for the study

AI.7. Monthly I Dcome

Work Typology
DW+SHG
DW+SHG+ME
DW+ML+SHG
DW+ML+ SHG+ME

Average monthly income

, of women (Rs.)
0
690
1113
1850

Average monthly family

income (Rs.)
2830
2942
1897
3110

The table above indicates that the income women get from micro-enterprises is
considerably less and in comparison to market labour (Rs. 690 and Rs.1113

167

respectively). Whereas the total family income of both the categories of women do
not differ considerably. It shows that women who are having higher family income
tend to remain at home.

168

Annexure II
Case I
Radha is aged 28, belongs to Hindu, Nair community and studied up to 8th standard.
She is the youngest daughter in the family having two married brothers. She works
as a made servant in a house 2 klns away from her house. She lives in a rented house
near her natal family where her mother stays. Her husband work as a mason in the
construction filed. She has two daughters aged 6 and 3. The first child goes to
school and the second one is attending anganwadi.
Her husband gives Rs.50 whenever he goes for work and it is one fourth of his daily
wage. He drinks excessively and she believes that he spends the remaining money
for it. She is the one managing household expenses. If there is short of money, she
has to arrange it by borrowing from NHG members or other relatives.
She is a member. of kudumbashree and attends the entire weekly meeting on
Saturdays.
She gets up early morning by 4AM. Collects water from a distant well and cook food
for children and keep them ready by 7.30 A.M. send them to her mothers place and
she goes for work. She has to 'start work by 8 AM. Children need to go to school
only by 9.30. She completes the work by 3 PM and while going back home, collects
the first child. Her mother collects the second by noon and takes care till she comes
back from work after collecting the first child. After reaching home, she takes the
second child along with her to a distant public well and to wash all the clothes.
She continues to work as house made because children get special food, even dresses
including uniform dress, there is a social custom if some one visits they pay tips to
servants, which is an extra income. However, she considers it as a low status work
and intends to shift to micro enterprises production along with other neighboring
women.
She managed to get loans under Ashraya programme, a housing scheme of
government of Kerala for the P?or. Bought a piece of land by selling gold and loan
taken from kudumbashree.
She has narrated her experience of negotiating with husband over purchasing the
land and building the house, which he was not very particular of. Since she made the
decision, had to take up the responsibility of the same, primarily mobilizing money
for the same. She believes that kudumbashree gives her.

169

Case 2
Lilly, 30 years old, studied up to plus 2, belongs to Ezhava community stays in her
husband's family with husband, mother in law and unmarried sister in law. She has
one daughter aged four years attending anganwadi. Her sister in law helps her in
sending and collecting the child_from anganwadi in her absence.
She is the president of the local ADS. She is with the micro enterprises producing
umbrella. It is a group-based activity with 5 more members from her neighourhood
group. She buys ready-made dresses from the nearby city and sells them to the
women from SHGs and other neighbouring houses on installment basis. Sister in law
helps. She is a major earning member in the family and takes major decisions. Her
husband is currently not working. Was working as a house painter.
Aspiring to become an active member of left party CPI (M). She got party
membership recently and wish to contest in the panchayat elections in future. She
attends party meeting and programmes in the nearby town. It gives more power and
respectability both within the sOciety and the SHG. Political aspirations are stronger
in her case and she identifies it as a major power center in the local political scenario.
She has narrated the instances of left party interference in panchayat functioning and
the kudumbashree affairs. Initially she was not happy with it, as she understood that
SHGs are primarily non-party development groups. Later on, she realized the
significance of the same and took party membership. She is involved in the
development project preparation under WCP in the pachayats and in all major social
functions.
As a result of the participation in Kudumbashree, she says that her communication
skills and leadership abilities have enhanced considerably. Also it gives her a
confidence to deal issues within the household too. She sees the enhanced mobility
as a positive outcome of her involvement in kudumbashree.

170

ease 3
Shoukath, belongs to Muslim community, aged 28 yrs, 10th class failed living in her
husbands family. There are 7 members in her family including two young brothers of
her husband. She needs to cook meals for the entire family and pack food for
husband's brothers who are going for work in the morning. Also, she needs to take
care of her three children. She has three children and they are of 2, 4 and 9 years old.
Since her mother in law is sick, she does all domestic work including caring her
mother in law. Her father in law is old and often falls sick. Hence the domestic
responsibilities are more and there is no other supporting hand in the household. She
says that though she is available at home, not in a position to give enough time for
her children. Her elder daughter helps her in taking care of young children.
She wish to engage in micro enterprise production unit with other members of the
neighbourhood group in future, once she gets relieved of the current domestic
responsibilities. Her husband doesn't like her to go out for work. He believes that it
shows his inability to meet family expenses being a male member of the household.
She is a member of Kudumbasbree, but husband doesn't like her attend the meeting
and hence she doesn't attend often. Her husband works as an electrician in a private
company in the nearby town. He gets Rs. 2000 monthly. Though she says that she
doesn't do work, she is engaged in cattle rearing and does stitching for the children
and neighbours. In the name of that she goes out and attend kudumbashree meetings.
She believes that there are many benefits from the government to the poor allotting
through SHGs and hence it is important to attend the meetings. Also, she sees its
value in terms of the friends she gets

C:0m kudumbashree SHGs. Since she is often

tensed and tired of domestic work, sees SHG meetings as an occasion to relieve
herself from the tensions and to ease.
She is sending her first child t? the local English medium school and sees it as an
important aspect of quality upbringing and intending to send the second child too in
the coming year.

171

Case 4
Veena, aged 27, belongs to dalit community and studied up to 7th standard.

She

lives in the housing colony for dalits with her husband. Her husband is a general
wage labourer. She has mentioned that the basic facilities such as drinking water,
toilets etc. are relatively poor in the colonies. She has to fetch water from a distant
well and to wash clothes there. As a result, there is more workload.
She works in the nearby town 'as a helper in a printing press. She gets Rs.1500
monthly. She spends Rs. 15-20 every day for children's food expense and gets eggs
or bakery sweets. She even buys books for the child read every month. She is a
member of kudumbashree and attends the meetings on Saturdays. But she is not very
active and made a remark that the caste status is important in NHGs and upper-caste
women dominate.
She has two children. Her son is 4 years old and attending anganwadi and the
daughter is 7 years old studying in the 2nd standard. Her neighbours and another
distant relative send her son in the morning and collect him back in the evening from
anganwadi. Her daughter goes on her own. She says that she gets good support from
her neighbours in the colony and. some of them are her relatives.
Her husband is a cook in a restaurant in the panchayat and he gets Rs.60 everyday in
addition to two meals. Veena hands over her salary to her husband and he is meeting
daily expenses. Nevertheless, in case of emergency, she has to borrow from
neighbours and from local moneylenders. She says that moneylenders prefer to lend
money to women because women repay promptly.

172

CaseS
Viji is 32 years old and passed standard 8th belongs to Hindu, vellala community. Her
daughter is 6 yrs old and studying in Ist standard and her son is 2 years old.
She works in the packing section of a flour and spices production and exporting unit
as a daily wage worker in the nearby town and leaves home at 9.30 AM. She gets
Rs.70 per day. She attends weekly meetings of SHG on Saturdays.
Her mother in law is a part-time made servant and she is free during 9AM to I PM
and takes care of the second child up to noon and then sends the child to Viji's
house. Her mother works as a part time sweeper in a warehouse and finishes work by
12 PM. Her mother collects the first child from the school. After her work, Viji goes
to her mother's house and collects both the children and get back to her in law's
house. She wishes to engage in micro-enterprise production and it is mainly due to
the flexible time arrangements of it.
Her husband is a daily wage worker and is currently unemployed. She says that he is
not very particular of finding jobs as she and her mother in law are earning income
and meet family expenses. There is frequent fight between she and her husband over
his alcohol consumption habit and he often asks money for the same. At times, it
ends up in she gets beaten up by him. She gets her wages on daily basis and does not
bring any money to home and gives it to a nearby daily collection center as a saving
before reaching home.
She sees kudumbashree as place where she can relax and talk to friends who
understand her problems. When she got pregnant third time, her friend in the NHG
accompanied her to hospital and terminated the pregnancy. She has not revealed the
matter to her husband, but informed her mother in law.

173

Case6
Rema, 26 belongs to Hindu Nair community and passed 12th standard and a
computer certificate course. Her daughter is 3 years old and attending anganwadi.
Her mother in law takes care of the child when she goes for work. She is planning to
send her daughter to an English medium school and to provide her high quality
education at a maximum possible level. Hence, she has decided to limit the number
of children in one.
She works as an assistant in a- .computer center in the town. She does typing job
works there. Rema leaves home at 9 AM and comes back at 4.30 PM. She gets
Rs.J200 monthly salary. She also manages a srD booth near her home from 6 PM
to IIPM and gets Rs.7S0 monthly. During this time, her child will be with her and
she teaches her alphabets and rhymes.
Her husband is a mechanic in a workshop and studied up to 8th standard. He leaves
for work by early morning and reached back home only in the late night. He gets
Rs.3000 monthly salary. He discusses household matters with Rema and considers
her opinion.
Rema is not very interested in actively participating in Kudumbashree groups. She
believes that it is meant for very poor, less educated women. Since she is working
with a computer center, she has many friends who belong to middle class, and
prefers to interact with them rather than with Kudumbashree members. However, she
has thrift savings in the group and takes thrift loans. She sees it mainly as a financial
instrument of the State. She could avail the benefit of computer education for free of
cost through Kudumbashree last year, since she is more educated than other
members.

174 .

Case 7
Jaya is 26 years old, belongs to Hindu, Nadar community and educated up to 5th
standard.
Her husband works as a general wage labourer and he has approximately 10 days of
work in a month. He gives Rs.1 00 to mother in law whenever he works and it is less
than half of his daily wage.
Jaya's daughter is 10 months old. She buys granite stones with few others and breaks
it into small pieces and sells th~m to a contractor every week. The place of work is
near her home and there is a granite quarry nearby. Her mother in law and sister in
law are taking care of the child. In emergency situations, however, she attends the
child since her place of work is near home. She has narrated that when the child is
sick, she takes off from her work and she compensates it by doing the work on
Sundays, the day normally she does not work. She earns Rs.250 weekly from this
activity.
Also, she sells vegetables in the local area and there are regular customers. She goes
early morning 5 AM to the city market and buys vegetables. She takes 3 hours in the
morning for this work and finishes it by 9.30 IS. She gets Rs.50 daily from selling
vegetables it is only during week days. She has explained that the intensity of work
at home and outside is giving her severe stress and hopes that once the child goes to
school, it may come down and she can concentrate more on her income generation
work. Though there are relatives to take care children and she is earning for the
entire family, the childcare given by substitute care providers is viewed in the family
as a help to her and not as their responsibility as well. She has clearly mentioned the
need for an institutional care arrangement in her locality with affordable cost.
She attends Kudumbashree meetings and wishes to join in micro-enterprise
production with other members. The work of breaking stones is having severe
adverse impact on her health and hence planning to leave that gradually. However,
she foresees the uncertainty in income that accrue from micro-enterprises.

175

Case 8
Sudha is 33 years and studied up to 11h standard. She belongs to Hindu, Nair
community. She has two children, son is 1 years and daughter is 4 years. The elder
one is studying in 2nd standard and the younger one is attending anganwadi. She lives
in a nuclear family and her in laws are living with her husband's brother.
Her younger son lives with her parents, who are staying 2 Ians. and she takes him
home in the week ends since she has to work in income earning activities during
week days and there is not other supporting hand.
Her husband is a tailor and run a tailoring shop in the panchayat in a rented room.
His income is seasonal and as the ready-made shops are coming up, it is gradually
declining too.
Sudha is engaged in a micro-enterprises unit that extracts coconut oil and sells. Also,
she lends money at higher rate of interest in her locality. She has taken Rs.l 0000 out
of her dowry money for the same and given the rest to her husband. She also works
as part-time sweeper in a private school. She says that the prevalence of dowry is
high and marriages are primarily decided based on the same. Her dowry was Rs.
25000 and 15 sovereigns (120 grams) of gold ornaments. She also owns 1 cents of

land and a small house that received as part of dowry.

Sudha has narrated instances of decision-making process in occasions where she and
,.

her husband had major disagreement. When she asked Rs.lOoo0 from her dowry
money to start money-lending business, her husband disagreed and he wanted to
invest the money to develop his shop. She insisted with a view that she has right over
that money and finally he has agreed. More over, the income from money lending is
substantially high than any other investment. Now she controls that income and
manages. She also wanted to give 5 sovereigns of gold to her younger sister as a gift
for her marriage and her husband objected. There was fight between her and her
husband took place over this issue and finally she has succeeded.

116

Case 9
Anie is 2S years old and passed 10th standard belongs to Christian, Nadar
community.
Her husband is an auto rickshaw driver and they live in their own house. The land, 4
cents, she got as her dowry and the house was contrasted by availing bank loan. She
has a son aged 4 years and attending kindergarten in an English medium school, 3
kms. away from home.
Anie runs a petty shop, which is a part of her house. It enables her to take care of
household and childcare works. She earns monthly Rs.2000 from the same. She
insists her husband to pay her Rs.70 every day from his income and she has saves
that amount. She meets daily household expenses.
She has mentioned that though the school fee for her son is high, it will bring many
benefits in future. Moreover, she has decided to limit number of children to one and
using contraceptive methods. There are two reasons that she has cited for it. One is to
provide better quality care to her son and another is that she needs to find an
additional income generating work, which requires more of her time and attention.
Aspiration for middle class social status was evident from her narrative on future
plans. She wants to renovate house and to buy modem consumer goods. Though she
is a member of Kudumbashree, she is not finding it as major avenue for income
generation. She has pointed th!\t the incomes are lower for women in many microenterprise production units in her locality. Again, she says that since is a space 'only'
for poor women, it does not give her much social exposure and interaction that she
gets now while running the petty shop.

177

Case 10
Leela is aged 29, belongs to Hindu, Nadar community and passed 4th standard. She
has two children and are of age

i years and 5 years.

She became orphaned at a young age and lived with her maternal uncle. Since she is
from poor background and there was no money as dowry, her uncle got her married
to a lorry driver who was already married and separated. She was living in a rented
house with her husband in the initial years of her marriage. Her husband used to
drink heavily and beat her up often. Since there was no money to survive, she started
working as house made in a neighbouring house. She could not continue her work, as
the salary was not regular. She has joined in Kudumbashree in 2001 and working as
an active member since then.
When the violence from her husband became unbearable and his support to the
family was stopped she has dec;:ided to separate from him. Also, he used to collect
her salary from the house she works as made servant. Since he resisted the separation
and threatened her, she filed a police complaint with the help of other members of
the SHG. Finally, he had to agree on it and stopped visiting her.
She says now that it is better not to have a husband than having a problem creating
one. She feels that she is more confident now in life and attributes its credit to
Kudumbashree and the members. She can make autonomous in decision-making, not
constrained. She bought 2 cents of land by both availing loan from SHG and using
the saving from post office. Also she got financial assistance from panchayat under
ashraya scheme, a housing scheme for the poor, and living there with her children.
Leela is engaged in many income-fetching activities. She is a fish vendor and starts
her work early in the morning all.d finishes the same by lOAM. She eams Rs. 70-1 ()()
daily from this work. In addition, she works in a marine export unit where dried fish
is exporting. Also, she does washing c10tJies for a nearby hospital in every alternate
day. Her total income from all the above work comes around Rs.3500 and she saves
Rs.l 000 every month.
When she goes for work, her children are care of by a girl in her neighbourhood and
she pays her Rs.400 monthly. She also pointed out the need for a proper childcare
institutional arrangement in the nearby area.

178

Bibliography
Aziz, Abdul (ed. ).1994. Poverl)l Alleviation in India - Policies and Programmes.
Ashish Publishing House, New Delhi, 1994.
Anker, Richard and Martha Anker.1995. "Measuring Female Labour Force with
Emphasis on Egypt", pp.l48-77

in Nabil F Khoury and Valentine Moghadam

(eds.), Gender and Development in the Arab World, London, Zed books.
Basu, A.M. and Basu, K. 1991. "Women's Economic Role and Child Survival: The
Case ofIndia", Health Transition Review, 1(1), pp. 83-103.
Basu, A.M. 1992. Culture. the status of Women.and Demographic Behaviour:
lllus/rated with the case ofIndia, Clarendon press, Oxford

Badgett, Lee M. V. and Nancy' Folbre. 2002. "Assigning care: Gender norms and
economic outcomes" pp. 327-44 in Marth Fethorolf Loutfi (ed.) Women, gender and
Work - what is equality and what do we get there? ILO, Geneva, Rawat

Publications, Jaipur.
Batliwala, Srilatha. 1994. "The Meaning of Women's Empowerment: New Concepts
from Action", in Gila Sen (ed.) Population Policies Reconsidered: Health.
Empowerment and Rights, Harvard University Press, Cambridge MA.

Becker, G.S. 1981.A Treatise on thefami/y. Harward University Press, UJ(

Beneria, Lourdes. 1982. "Accounting for women's work", in Lourdes Beneria (ed.)
Women and Development: The'Sexual Division of Labour in Developing Societies,

New York, Praeger, pp. 119-47.

- - -.. 1992. "Accounting for Women's Work: Progress of Two Decades", World
Development. 1(22), pp.I 547-60.

___ . 1995. "Towards a greater integration of gender in economics", World

Development, 23( II )pp.

179

Beneria, Lourdes and Martha roldan. 1987. The Crossroads of Class and Gender:
Indistrial Homework. Subcontracting and Household dynamics in Mexico City,
Chicago, University of Chicago Press.
Beenstock, M and Sturdy, P. 1990. "The Detenninants of Infant Mortality in Regional
India", World Development, Vol. 18(3), pp.443-453.
Bhat P.N.M. and S. Irudaya Rajan. 1990. "Demographic transition in Kerala
Revisited", Economic and Political Weekly, September 1-8, pp.1957-1979.
Bianchi, S M. 2000. "Maternal Employment and Time with Children: Dramatic Change
01' Surprising Continuity?", Demography, 37, ppAOI-14

Boserup, Ester. 1970.

Women's Role in Economic Development, New Yark, St.

martin's Press.
Bubeck, Elisabeth Diemut.1995. Care, Gender and Justice, Oxford, Clarendon

Press.
Chathukulam, Jose and M.s. John. 2002. "Five Years of Participatory Planning in
Kerala: Rhetoric and Reality", Economic and Political Weekly, 7 December.
Cadwell J.C. 1979. "Education as a Factor in Mortality Decline: An Examination of
Nigerian Data", Population Studies, 33, pp.395- 413.
---.1986. "Routes to low mortality in poor countries", Population and
Development Review, 12, pp.171-220.
---.1997. "The global fertility transition: The need for a unifying theory",
Population and Development Review,23, 803-\2.
Cadwell J.C and P. Calwell. 1993."Women's Position and Child Mortality and
Morbidity in Less Developed Countries pp.122-139, in N Federici, K.O.Mason and S
Sogner (eds.) Women's Position and Demographic Change, Oxford, Clarendon Press

180

- - . "How is greater Maternal Education Translated into Lower Child Mortality?",

Health Transition Review, 4 pp.224-29

Christensen Kathleen. 1988. Women and home-based work: The unspoken contract
New York, Henry holt.
Cloud, Kathleen and Garrett, Nancy (1996), "A Modest Proposal for Inclusion of
Women's Household Production in Analysis of Structural Transformation", Feminist

Economics, 2(3), pp.93-120

Das Gupta, M. 1987. "Selective Discrimination against Female children

In

Rural

Punjab, India", Population Development Review, VoU3 (I), pp.77-IOO.

Desai Sonalde and Devaki Jain. 1994. "Maternal Employment and Changes in Family
Dynamics: the social context of Women's work in rural south India", Population

Development Review, N0.20, pp. 115-133.

Dietrich, G. and N. Nayak. 2002. Transition or Transformation? A Study of the

Mobilisation, Organisation a~d the Emergence af Consciousness among the

Fishworlcers ofKerala, India, Madurai: Tamil Nadu Theological Seminary.
Dixon, Ruth B.1978. Rural Women at Work: Strategies for Development in South

Asia, John Hopkins University Press, Baltimore.

- - . 1982. "Women in Agriculture: Counting the Labour force in Developing
Countries", Population and Development Review, 8(3), pp.539-66
Donahoe Anne Debra. 1999. "Measuring Women's Work in Developing countries"

Population Development Review, 25(3), pp.~43-76.

Dre'ze Jean and Mamtha Munhy. 1999. "Female Literacy and Fertility: Recent
Census Evidence from India", mimeograph, Centre for History and Economics,
King's College, Cambridge, U.K.

181

Dyson, Tim and Mick Moore. 1983. "Kinship Structure, Female Autonomy and
Demographic Behaviour in India", Population and Development Review, 9( I), 35-60
Eapen, Mridul .1994. "Rural Non-agricultural Employment in Kerala: Some
Emerging Tendencies", Economic and Political Weekly, Oct.26
Eapen, M. and Soya Thomas .2005. "Gender Analysis of Select Gram (Village)
Panchayats Plan-Budgets in Tri~andrum District, Kerala", Discussion Paper No. 11:
HDRC, UNDP, New Delhi.

Eapen, Mridul and Praveena Kodoth. 2002. "Family Structure, Women's Education
and Work: Re-examining the High Status of Women in Kerala", Working Paper No.
341. Centre for Development Studies, Thiruvananthapuram, Kerala.
Engels, F. 1972. The Origin ofthe Family Private Property and the State, path finder,
New York
England, P. 2000. "Conceptualising Women's Empowerment in Countries of North" in
Presser,H.B, and Sen G (eds.),Women's Empowerment and Demographic Processes:
Moving beyond Cairo, Oxford U~iversity press, Oxford

Erwer, Monica. 2003. Challenging the Gender Paradox: Women's Collective Agency
and the Transformation of Kerala Politics, Goteborg: Department of Peace and

Development Research, Goteborg University.

Floro, Maria S. 1995. "Economic Restructuring, Gender and Allocation of Time",
World Development, 23(11), pp. 9\3-29

Floro, Maria S and Gary Dymski . 2000. "Financial Crises, Gender and Power: An
Analytical Framework", World Development, 28(7) pp.1269-83
Folbre, Nancy. 1983. "Of PatFiarchy born: The Political Economy of Fertility
Decisions", Feminist Studies, 9, pp.26 1-84.

182

--.1995. "Holding Hands at Midnight: The paradox of Caring labour", Feminist

Economics 1(1), pp.73-92

- - . 1994. Who pays for the Kids? Gender and the Structures of Conseraint, New
York, Routledge.
Folbre Nancy, Jayoung Yoon, kade finoffand Allison sidle Fuligni .2005. "By What
Measure? Family Time Devoted to Children in the United States", Demography, 42
(2) pp.373-390
Fraser Nancy and Linda Nicholson. 1990. "Social Criticism without Philosophy: An
encounter between feminism and post modernism" in Linda Nicholson (ed.),

Feminism/Post Modernism pp.19-38, New York, Routledge.

Gallo, Ester. 200S. "Unorthodox Sisters: Gender Relations and Generational Change

in Malayalee Transnational Marriages", Paper presented at KCHR colloquium,

Trivandrum, February S.
Gillepsie Stuart and Mc Neil Geraldine. 1992. Food Health and Survival in India

and Developing couneries, Oxford University Press, London

GilIigon Caro1.1982. In a different Voice:Psychological Theory and Women's

development, Cwnbridge, Harward University Press.

Greenhalgh Susan.199S. "Engendering population policy and practice in rural China:
For a femnist demography of reproduction", Signs, 20, pp.60 1-41.
Gupta, Jyotsna Agnihothri. 2000. New Reproductive Technologies, Women' Health

and Autonomy, Sage Publications, New Delhi.

Hartmann Heidi. 198\. "The Family as the locus of Gender, Class and Political
Struggle: The Exwnple of Housework", Signs, 6, 366-94.

183

Hertz, R. 1999. "Working to Place Family at the Center of Life: Dual-eamer and
single-parent strategies", Annals of tire American Academy of Political and Social

Science. 562, pp.l6-31.

Hill, S.M. 1985. "Patterns of Time Use", pp.\33-76 in Juster F.T and Stafford F.P
(eds.), Times Gaads and Well-being. Ann Arbor, MI, Survey Research Centre,
University of Michigan.
Himmelweit Susan and Mohun Simon.1977. "Domestic Labour and Capital",

Cambridge Journal ofEconomics, I, pp.15-31.

Hobson barabara and Helena Bergman. 2002. "Compulsory Fatherhood: The
Coding of Fatherhood in the Swedish Welfare State", pp.94-124 in Barabara
Hobson (ed.), Making Men into Fathers: Men Masculinities and the Social Politics

ofFatlrerhaad, Cambridge, Cambridge University Press.

Hussmanns,

Ralf,

Farhad

Mehran

and

Vijay

Verma. 1990.

"Surveys of

Economically Active Population", Employment, Unemployment and Under

Employment', ILO Report. Geneva.
Issac,

Thomas

and

Richard

2000.

Franke

Local

Democracy

and

Development:People's Campaign for Decentralised Decentralised Planning in

Kerala. Left World, New Delhi
Isaac,Thomas T.M., Michelle Williams, Pinaki Chakraborty, Binitha V. Thampi .
2002. "Women Neighbourhood Groups: A New Perspective", Paper presented at tire

seminar

'Decentralisation.

Sustainable

Development and Social Security',

Alappuzha, May 11-12.

Jackson, R. M. 1998. Destined for Equality: The Inevitable Rise of Women's Status.
Cambridge, Mass, Harvard University.
Jeffrey, Robin. 2003. Politics. ,Women and Well Being: How Kerala Became a

"Model", New Delhi, Oxford University Press.

184

Jejeebhoy, Shireen J. 1995. Women's Education, Autonomy, and Reproductive

Behaviour: Experience from Developing Countries, Clarendon Press, Oxford.

Jejeebhoy, Shireen J. 2000. "Women's Autonomy, in rural India: Its Dimensions,
Determinants, and the Influence of Context" in Women's Empowerment and

Demographic Processes: Moving beyond Cairo (ed.) Presser,H.B, and Sen G, Oxford
University press, Oxford
Juster, F. Thomas. 1985. "The validity and quality of time use estimates obtained
from recall diaries", in Time, G.?ods, and Well-Being, edited by T. F. Juster and F.
Stafford, P. Ann Arbor, MI, Institute for Social Research.
Juster, F. T., and Stafford, F. P. 1985. Time, Goods, and Well-being, Ann Arbor MI,
Institute for Social Research, University of Michigan.
Kabeer, Naila. 1999. "Resources, Agency, Achievements: Reflections on the
Measurement of Women's Empowerment", Development and Change, 300, pp.435464

- - . 1994. Reversed realities: Gender hierarchies in development thought, New

Delhi, Kali for Women.
Kadiyala, Suneeta .2004. "Scaliiig up Kudumbashree - Collective Action for Poverty
Alleviation and Women's Empowerment", FCND Discussion Paper No. 180,
International Food Policy Research Institute, Washington.
Karlekar, Malavika.l982. Poverty and Women's Work: A study of Sweeper Women

in Delhi, Vikas Publishing House, New Delhi.

Khan Osmani LN. 1998. "Impact of Credit on the Relative Well-Being of Women:
Evidence from the Grameen Bank", IDS Bulletin, 29(4), Sussex.
Kishore, Sunitha.l993. "May God give sons to all: Gender and Child mortality in
India", American Sociological Review, 58(2), 247-65.

18S

-.1992. "All 'Devis' but not all wanted: A district level analysis of female
discrimination in India, 1961-81", Paper presented at the 1992 annual meeting of the

Population Association ofAmerica, Denver, Colorado.

Krishnaji, N. 1995. "Working Mothers and Child Survival in Rural India: Insights from
spacial pattern", &onomic and Political Weekly, November, pp. 2803-2808.
Lindberg, Anna.200 I. Experience and Identity: A Historical Account of Class. Caste

and Gender among the Cashew Workers of Kerala. 1930-2000. Lund: Department of
History, University of Lund.
Losano Beverly. 1989. The Invi;ible Workforce: Transforming Business with Outside

and Home-based Workers. New York, free Press.

Lundberg, Shelly & Pollak, Robert A. 1997. "Bargaining and distribution in
marriage", in Persson, I. and Jonung C (Eds.), Economics of the family and family

Policies. Routledge, London and New York.

Mason, Karan

Oppenheim.

1993. "The impact of Women's Position on

Demographic Change during the Course of Development" in Nora Federic (ed.)

Women's Position and Demographic Change, Clarendon Press, London

Mathew, E T .1995. "Educational Unemployment in Kerala: Some Socio-Economic
Aspects", &onomic and Politic~l Weekly, 30(6).

---.1997. Employment and Unemployment in Kerala. Sage Publications, New

Delhi.
Mayoux, Linda 1998, "Micro-finance Programmes and Women's Empowerment:
Approaches, Evidence and the Ways Forward", Discussion Paper. Open University,
Milton Keynes.
--.

1999. "Questioning Virtuous Spirals: Micro-finance and

Women's

Empowerment in Africa", Journal ofInternational Development, 11.

186

----. 1998. "Participatory Learning for Women's Empowerment in Micro-Finance

Programmes: Negotiating Complexity, Conflict and Change", IDS Bulletin. 29 (4),
Sussex.
Mencher, J.P. 1988. "Women's work and Poverty: Women's Contribution to Household
Maintenance in South India" in Dwyer, D and Bruce (eds.), A Home Divided, Stanford
University Press, California.
Mukherjee C and Thomas Isaac, T M.1994. "Nine Observations on Educated
Unemployment in Kerala", Background paper presented at a one-day workshop at

Centre for Development Studies, January 17.

Murthy, Mamtha, Anne-Catherine Guio and Jean Dre'z .199S. "Mortality, Fertility
and Gender Bias in India: A District Level Analysis", Population Development

Review. 21(4) pp.74S-782.

Nair, N.D.Gopinathan. 2000. "People's Planning in Kerala: A Case Study of Two
Village Panchayats", Discussion Paper No.6. KRPLLD, Thiruvananthapuram.
Oommen, M A. 1992. "The Acute Unemployment Problem of Kerala: Some
Exploratory Hypothesis", lASSI Quarterly. 10(3).
- - - . 1993. Essays on Kerala Economy. Oxford &IBH publishing co. Pvt. Ltd.,
New Delhi.
-------.1999. The Community Development Society of Kerala: An Impact Study Vol.
I, Institute of Social Sciences. New Delhi
----. 200S. 'Deepening Decentralised Governance in Rural India: Lessons from
the People's Plan Initiative of Kerala', Sociai Change and Development, 3.
Osella, Filippo and Caroline Osella.2000. Social Mobility in Kerala: Modernity and

Identity in Conflict, Pluto Press, London.

187

Papanek, Hannah .1979. "Famil~ Status Production: The 'Work' and 'Non-Work' of
Women", Signs 4(4).
Parcel, T.L and Menaghan, E.G. 1994. "Early Parental Work, Family Social Capital
and Early Childhood Outcomes", American Journal ofSociology, 99, pp.972-1 009
Perry-Jenkins, Repetti, R. and Crouter, A. 2000. "Work and Family in the I990s",

Journal ofMarriage and the Family, 62(4), pp.98 1-98.

Pleck. J.H .1985. Working WifeslWorking husbands. Beverly Hills, CA, Sage
Publications.
Prakash, B A.1988. "Educated Unemployment in Kerala: Some Observations Based
on a Field Study", working paper. 224, Centre for

Development Studies,

Thiruvananthapuram.
Presser Harriet B. I 997."Demography, Feminism and Science-Policy Nexus",

Population and Development Review, 23(2), pp.295-33I.

Rajasekhar D. 1996. "Problems and Prospects of Group Lending in NGO Credit
Programmes in India", Savings and Development, Quarterly Review, 10 (I) pp.79-

104.
Ramasubban, Radhika and Jejeebhoy, Shireen J (eds.) 2000. Women's Reproductive

Health in India, Rawat Publications, New Delhi.

Rhyne, E. and M. Otero (1994) "Financial Services for Micro-enterprises: Principles
and Institutions" in Otero, M and Rhyne,E eds The New World of Microenterprise

Finance: Building Healthy Financial Institutions for the Poor, IT Publications,

London.
Riley N. E. 1999. "Challenging Demography: Contributions from Feminist theory",
Sociological Forum, 14(3), pp.369-97.

188

Robinson, J. P. 1985. "The validity and reliability of diaries versus alternative time
use measures", In Juster, T. F. -and Stafford, F. (Eds.) Time, goods, and wellbeing.
Ann Arbor, MI: Institute for Social Research, University of Michigan.
Robinson, John and M. Milkie. 1998. "Back to the Basics: Trends in and Role
Determinants of Women's Attitudes toward Housework", Journal of Marriage and
the Family 60, pp.205-18.

Sanchez, L anf Thompson, E. 1997. "Becoming Mothers and Fathers: Parenthood,

Gender and Division of Labour", Gender and Society, 11, pp.747-72.
Seema, T.N.2001. "Performance of Anganwadi centres in Kerala: An evaluation and
experiment to develop a model centre with community participation".Discussion

Paper No. 28, KRPLLD, Thiruvl\llanthapuram.

Sen ,Geeta and Batliwala, Sreelata. 2000. "Empowering Women for Reproductive
Rights" in Women's Empowerment and Demographic Processes: Moving beyond

Cairo (ed.) Presser,H.B, and Sen G, Oxford University press, Oxford

Sen, Amarthya, 1990. "Gender and Co-operative Conflicts", in Irene Tinker (ed.)

Persistent Inequalities: Women and World Development, New York, Oxford

University Press.

--.2000. Development as Freedom, New Delhi, Oxford University Press.

Short E Susan, Feinian Chen, Barbara Entwisle and Zhai Fengying. 2002. "Maternal
work and Childcare in China: A'multi-method analysis", Population and Development
Review, 28(1), pp.3I-57.

Sivakarni, M. 1997. "Female Work Participation and Child Health: An investigation in

Rural Tamil Nadu", India. Health Transition Review, No.7, pp. 21-32.
--.2000. "Does maternal Employment Influence Child Care and Morbidity?
Evidence from Poor population, Demography India, 9( 1).

189

Stancey Judith and Barrie Thome. 1985. "The missing feminist revolution in
sociology", Social Problems, 32, pp.301-16.
Tharakan, P.K. Michael. 2004. "Historical Hurdles in the Course of the People's
Planning Campaign in Kerala, India", in John Harris, Kristian Stokke, Olle Tornquist
(Eds.) Politicising Democracy: Local Politics and Democratisation in Developing

Countries, London: Pal grave Macmillan.

Tornquist, Olle. 2004. "The Political Deficit of Substantial Democratisation", in
John Harris, Kristian Stokke, Olle Tornquist (Eds) Politicising Democracy: Local

Politics and Democratisation in Developing Countries,

London:

Palgrave

Macmillan.
Thampi, Binitha. V. 2001. "Community Development Society in Alleppey: An
Evaluation

Study",

Evaluation

Study

Report,

State

Planning

Board,

Thiruvananthapuram.
Thompson L and Walker, A.J. 1990. "Gender in Families: Women and Men in
Marriage, Work and Parenthood", Journal ofMarriage and the Family, 51, pp.845-71.
Tinker, Irene. 1990. "The Making of a Field: Advocates, Practitioners and Scholars" in
Irene Tinker (ed.), Persistent Inequalities: Women and Development, New York,
Oxford University Press, pp.27-53.
United Nations, 1985. Socio-economic Differential in Child Mortality in Developing

Countries, Department oflnternational Economics and Social Affairs, New York.

Unni, G. 1989. "Change in

E~nomics Employment in Rural Areas, 1%1-1983",

Economic and Political Weekly, April 29, pp..WS. 23-WS 31.

Visaria, Leela. 1996. "Regional Variations in Female Autonomy and Fertility and
Contraception in India", in Jeffrey and Basu (ed.) Girls Schooling, Women's

Autonomy and Fertility Change in South Asia, Sage Publications, New Delhi.

190

Ware, Helen. 1984. "Effects of Maternal Education, Women's Roles and Child Care on
Child Mortality", Population and Development 10, pp.l91-214.
Waldfogel. 2002. "Childcare, Women's Employment and Child Outcomes", Joumal of
Population Economics, pp. 15(3): 527-48
Watkins, Susan. 1993. "If all we knew about women were what we read in
Demography, what would we know?", Demography, 30, 551-78.
Zachariah K.C. et.al. 1994, Demographic Transition in Kerala in the 1980s, Centre for
Development Studies Monograph Series, Trivandrum.

191