STUDY OF WOMEN’S EMPOWERMENT
IN THE DISTRICT OF BANKURA
SUBMITTED BY
PAPITA DUTTA
TO
THE UNIVERSITY OF BURDWAN
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN
ECONOMICS
2014
UNDER THE GUIDANCE AND SUPERVISION OF
DR. MANIKLAL ADHIKARY
PROFESSOR OF ECONOMICS
BURDWAN UNIVERSITY
Dr. Maniklal Adhikary
Professor
Department of Economics
The University of Burdwan
Golapbag, Burdwan-713104
West Bengal (INDIA)
Dial (0342)-2558-554, Extn. 438 (O)
+91-9476-233029
+91-9434-660220
Date:
Certified, that Smt. Papita Dutta, a registered scholar in Economics, Burdwan
University, Golapbag, Burdwan, West Bengal, has duly completed her research work
under my supervision. She is keen on submitting her thesis entitled “Study of Women’s
Empowerment in the District of Bankura”. I have approved her thesis and permitted her
to submit it to the University of Burdwan for Ph. D. degree in Economics. Further,
certified that neither this dissertation nor any part thereof was submitted to this or any
other university in the country or abroad for Ph. D. or any other degree. However, it may
also be noted that Smt. Dutta has delivered two seminar lectures on her research work in
partial fulfillment of the requirement for the submission of the Ph. D. thesis and compiled
with all relevant conditions specified in the resolutions of the University of Burdwan.
Maniklal Adhikary
i
Preface
The importance of human capital juxtaposed with physical capital has been emerging
since the end of the Second World War to pace economic development in the developing
countries. Empowerment is no doubt the pivotal component of human capital.
Empowerment of women has emerged as a developmental issue after the inception of the
concept of ‘capability approach’ for development. Before that, women’s empowerment
was the subject of the feminists and viewed as a socio-political issue. The feminists look
women’s empowerment entirely as an instinct matter. After 1980s economists have
recognized women’s empowerment as an instrument for human development. In addition
to the feminist goals the instrumental idea of women’s empowerment have some
important policy pay-offs. Advocates of instrumentalists have tried to define women’s
empowerment with a broader vision compared to the vision of feminists. However,
women’s empowerment has still not been clearly defined and segregated from other
closely related concepts like gender inequality. Women’s empowerment is viewed as
mechanism of improving the qualitative aspects of women. In this dissertation we have
studied women’s empowerment as means and ends of forming human capital.
Nowadays women’s empowerment is a multidimensional and multi-level concept. The
study of women’s empowerment is also context specific. We have premeditated
empowerment at the household level and at the community level for the women of
Bankura district. For this purpose we have considered five dimensions of women’s
empowerment for each level. These are economic dimension, familial dimension,
political dimension, social dimension and legal dimension. Further, each dimension has
been covered by some suitable indicators in the context of Bankura district. Combining
all the dimensions we have computed empowerment index for each sample woman at the
household level and at the community level. This empirical study is based on a set a
primary data collected from 580 women of Bankura district. An attempt has been taken
to estimate the impact of women’s empowerment at the household level and at the
community level on family planning decision, on the incidence of domestic violence
against women, and on spending for children’s education in Bankura district. The study
ii
has also analysed the nature and causes of the low level empowerment of the women in
Bankura district.
The corpus of this dissertation has been structured into six chapters. In chapter one, we
present the relevance and the objectives of this study. The second chapter deals with the
review of relevant literatures. Chapter three has explained the theoretical framework for
the estimation of the impact of women’s empowerment on the issues of household
welfare and for investigation of the factors affecting women’s empowerment. Empirical
models and hypotheses for empirical testing have also been formulated in this chapter.
Chapters four has analysed the components of empowerment of the women in Bankura
district, West Bengal. We have found that empowerment of the sample woman at the
household level and at the community level have not reached at the creditable level. We
have explained the empirical impact of empowerment variables along with other factors
on household and child welfare in chapter five. In this chapter the empirical estimates of
the women’s empowerment have also been explained. Our empirical findings shows that
women’s empowerment is instrumental in improving the probability of taking family
planning decision, in alleviating the incidence of domestic violence and in increasing the
share of household income spend for children’s education. Finally, we find age, personal
occupation, personal income, financial inclusion, household’s landholding, SHG-
membership, educational background of the female household members as crucial factor
for enhancing empowerment at the household level and at the community level. On the
basis of the empirical findings, we draw the policy prescription in chapter six for
improving household welfare and empowerment of the women in a better way.
I want to extend my thanks and gratitude to all of those who have supported me in the
process of preparing this dissertation. It is an output of collective efforts of many
individuals.
I particularly honour and thank Dr. Maniklal Adhikary, Professor of Economics,
Burdwan University, for his unwavering support and constant inspiration for preparing
this dissertation. He enriches this dissertation with his valuable advices and precious
commands. I am fortunate enough to have him as my teacher and supervisor. Without his
iii
guidance and support, it was quite impossible for me to proceed towards my dream of
preparing this Ph. D. dissertation.
My parents, Rina Dutta and Sisir Dutta need no gratitude for their unconditional love,
affection, patience and support that they have shown throughout my life. Actually, it is
their meticulous efforts that help me prepare this thesis. I am thankful to God for giving
me such loving and caring parents.
I am thankful to my beloved husband, Dr. Supravat Bagli, for encouraging me and
particularly for his support and cooperation during the course of data collection. He has
helped me enrich this dissertation with his valuable commands and suggestions. I am
thankful to my only son, Arya (Reetobrato Bagli) because he tolerates my all absence
with patience. Though he has been deprived of my love and affection during the period of
preparing this thesis, he understands me in his own little ways. I remain indebted to him for
this sacrifice in rest of my life.
I want to express my earnest thanks to my loving sister, Sangita Dutta, for her absolute
support. She accompanied my son during my hard time in preparing my thesis.
I am thankful to Dr. Madhuri Adhikary for providing me home like ambience. I would
like to thank Sri Ajoy Kumar Majhi and Smt. Chitra Sadhu, who sincerely supported me
during my fieldwork.
I am grateful to Professor Arup Kumar Chattopadhyay, Professor Soumeyendra Kishore
Dutta, Dr Atanu Sengupta, Dr. Pravat Kuri, Dr. Rajarshi Majumdar, Dr. Anindita Sen
and all other teachers of the Department of Economics, Burdwan University, who have
provided valuable suggestions and comments for enriching this dissertation.
I am thankful to the authorities of Burdwan University for giving me the opportunity to
execute my research works. I extend my heartfelt thanks to the staff members of the
department of Economics, for their invariable support. I extend my sincere appreciation
to the staff members of Indian Statistical Institute, Kolkata, and the staff members of
iv
Guskara Mahavidyalaya, Guskara who have supplied valuable documents, information
and reading materials related to my dissertation.
.
Last but not the least; I am ever thankful to all respondents and personalities who spend
time and effort for me during the time of household survey. There are too numerous
friends and relatives who remain unnamed here to receive my heartfelt appreciations.
I am humble to state that I have prepared this dissertation with the best of my knowledge
and efforts. Despite my sincere efforts in the preparation of this dissertation there may
have some mistakes for which I am solely responsible.
28th
May, 2014 PAPITA DUTTA
Burdwan
v
CONTENTS
List of Tables x
List of Figures xii
Chapter One 1-30
WOMEN’S EMPOWERMENT
AN INSTRUMENT FOR HUMAN DEVELOPMENT
1.1. Introduction 1
1.2. Idea of Women’s Empowerment 4
1.3. Women in India and in the District under Study: A Fact Sheet 9
1.4. Relevance of the Study 14
1.5. Justifications of the District Selection 16
1.6. Socio-Economic-Demographic Characteristics of Bankura District 17
1.6.1. Geographic and Administrative Profile 18
1.6.2. Demographic Characteristics 19
1.6.3. Socio-Economic Characteristics 21
1.6.4. Educational Status 22
1.6.5. Empowerment Statistics 24
1.7. Objectives of this Study 25
1.8. Conclusion 30
vi
Chapter Two 31-72
REVIEW OF LITERATURES ON WOMEN’S EMPOWERMENT
2.1. Introduction 31
2.2. Women’s Empowerment: Issues and the Conceptions 32
2.3. Women’s Empowerment and Welfare: The Impact Studies 41
2.4. Studies relating to Determinants of Women’s Empowerment 51
2.4.1. Studies on Women’s Empowerment: The Indian Scenario 51
2.4.2. Studies on Women’s Empowerment: The Global Scenario 61
2.5. Conclusion 72
Chapter Three 73-120
MODEL, METHODOLOGY AND DATA
3.1. Introduction 73
3.2. Measures of Women’s Empowerment 74
3.2.1. Selected Dimensions and Indicators of Women’s Empowerment 76
3.2.1A. Dimensions and Indicators of Women’s Empowerment at
Individual or Household level 76
3.2.1B.Dimensions and Indicators of Women’s Empowerment at Community level 78
3.2.2. Degree of Women’s Empowerment 79
3.2.3. Analytical Framework for Computing Composite Index of
Women’s Empowerment 80
3.3. Research Design for Studying the Impact of
Women’s empowerment on Household Welfare 82
3.3.1. Analytical Framework for Studying Impact of
Women’s Empowerment on Family Planning Decision 82
3.3.2 Analytical Framework for Impact of Empowerment on
Domestic Violence against Women 85
vii
3.3.3. Analytical Framework for Impact of Women’s Empowerment on
Children’s Educational Expenditure 87
3.4. Analytical Framework for Studying Women’s Empowerment 90
3.5. Regression Specification of the Analytical Models relating to
Women’s Empowerment 92
3.5.1. Probit Models for Decision regarding Family Planning 92
3.5.2. Logit Models for Incidence of Domestic Violence 93
3.5.3. Log-Lin Models for Children’s Education Expenditure as
Proportion to Household Income 93
3.5.4. Linear Regression Models for Women’s Empowerment at
the Household Level and at the Community Level 94
3.6. Definition and Measurement of the Variables included in
the Regression Models 96
3.7. Specification of Hypothesis 105
3.7.1. Hypotheses relating to the Model for Decision regarding Family Planning 105
3.7.2. Hypotheses relating to the Model for Incidence of Domestic Violence 108
3.7.3. Hypotheses in Connection with the Child Education Expenditure 111
3.7.4. Hypotheses relating to the Models of Women’s Empowerment 113
3.8. Methodology of Data Collection 115
3.8.1. Sampling Design 116
3.8.2. Profile of the Sample Areas 117
3.8.3. Nature and Scope of Data 118
3.8.4. Diagnostic Check for the Sample Size 119
3.9. Conclusion 120
Chapter Four 121-156
COMPONENT ANALYSIS OF WOMEN’S EMPOWERMENT
4.1. Introduction 121
4.2. Descriptive Statistics of the Surveyed Households 122
4.2.1. Categorical Characteristics of the Sample Households/Individuals 122
viii
4.2.2. Socio-Economic and Demographic Characteristics
of the Surveyed Population 127
4.3. Frequency Distribution of the Indicators of Women’s Empowerment
at Household Level and at Community Level 130
4.4. Analysis of Women’s Empowerment Indices 137
4.4.1. Outcomes of Principal Component Analysis
of the Household Level Empowerment of Women 137
4.4.2. Outcomes of Principal Component Analysis
of the Community Level Empowerment of Women 141
4.4.3. Descriptive Statistics of the Empowerment Indices 144
4.5. Bivariate Correlation among Selected Variables 147
4.6. Conclusion 155
Chapter Five 157-198
EMPIRICAL ESTIMATES AND ANALYSIS
5.1. Introduction 157
5.2. Impact of Women’s Empowerment on Decision regarding Family Planning 157
5.2.1. Model-1A: Probit Model with Simple Empowerment Indices 158
5.2.2. Model-1B: Probit Model with Composite Empowerment Indices 164
5.3. Impact of Women’s Empowerment on Domestic Violence against Women 167
5.3.1. Model 2A: Logit Model with Simple Empowerment Indices 168
5.3.2. Model-2B: Logit Model with Composite Empowerment Indices 174
5.4. Impact of Women’s empowerment on Expenditure for Child Education 176
5.5. Determinants of Women’s Empowerment in Bankura District 182
5.5.1. Determinants of Women’s Empowerment at the Household Level 183
5.5.2. Determinants of Women’s Empowerment at the Community Level 190
5.6. Conclusion 198
ix
Chapter Six 199-221
POLICY PRESCRIPTIONS
6.1. Introduction 199
6.2. Major Findings 199
6.3. Existing Policies and Programs towards Empowering Women 203
6.3.1. National Policy and Legislation for Women in India 203
6.3.2 Governmental Programmes for Enhancing Women’s Empowerment in India 206
6.4. Policy Prescriptions and Implications 210
6.4.1. Suggestions for Empowering Women in Bankura District 211
6.4.2. Suggestions for Improving the Likelihood towards Family Planning 214
6.4.3. Suggestions for Alleviating Domestic Violence against Women 216
6.4.4. Suggestions for Improving Child Education Expenditure 218
6.5. Conclusion 220
Bibliography 222-234
x
Tables
Table- 1.3.1 Sex Ratio: Number of Female per Thousand Male 10
Table-1.6.1 Geographical Area (Square Kilometres) 18
Table-1.6.2 Administrative Profile of Bankura District 19
Table-1.6.3 Population and Decadal Growth rate by residence 20
Table-1.6.4 Percentage Distribution of Rural Urban Population 20
Table-1.6.5 Sex Ratio in West Bengal and Bankura District 21
Table-1.6.6 Caste Wise Population (Percentage) with Reference to 2001 Census 21
Table –1.6.7 Human Development Index of Bankura District 21
Table-1.6.8 Percentage of Population According to Occupational Status in 2001 22
Table 1.6.9 Gender Wise Distribution of Literates 23
Table-1.6.10 Sex Wise Literacy Rate among Scheduled Caste, Scheduled Tribe 23
Table 3.8.1 Area Specific Distribution of the Sample Households 117
Table-4.2.1 Distribution of the Categorical Variables of the Sample Households 124
Table-4.2.2 Percentage Distribution of the Individual Categorical Variables 126
Table-4.2.3 Relevant Demographic Profile of the Sample Households 127
Table-4.2.4A Socio-economic Characteristics of the Sample Households 128
Table-4.2.4B Socio-economic Characteristics of the Sample Households 129
Table-4.3.1A Percentage Distribution of the Indicators of
Women’s Empowerment at Household Level 131
Table-4.3.1B Percentage Distribution of the Indicators of
Women’s Empowerment at Household Level 133
Table-4.3.2A Percentage Distribution of the Indicators of
Women’s Empowerment at the Community Level 135
Table-4.3.2B Percentage Distribution of the Indicators of
Women’s Empowerment at the Community Level 136
Table 4.4.1 Results of KMO and Bartlett’s Test for Sample Adequacy for
Factor Analysis of Women’s Empowerment at the Household Level 137
xi
Table-4.4.2 Total Variance in the Indicators of Women’s Empowerment
Explained by the Components at the Household Level 139
Table-4.4.3 Rotated Component Matrix of Women’s Empowerment
at Household Level 140
Table-4.4.4 Results of KMO and Bartlett’s Test for Sample Adequacy for
Factor Analysis of Women’s Empowerment at the Community Level 141
Table-4.4.5 Total Variance in the Indicators of Women’s Empowerment
Explained by the Components at the Community Level 142
Table-4.4.6 Rotated Component Matrix of Women’s Empowerment
at the Community Level 143
Table-4.4.7 Descriptive Statistics of the Women’s Empowerment 145
Table-4.4.8 Percentage Distribution of Women’s Empowerment in Bankura District 145
Table-4.5.1 Bivariate Correlation Matrix 148
Table-4.5.2 Bivariate Correlation Matrix 149
Table-4.5.3 Bivariate Correlation Matrix 151
Table-4.5.4 Bivariate Correlation Matrix 152
Table-4.5.5 Bivariate Correlation Matrix 154
Table-5.2.1 Results of the Probit Model for the Decision regarding Family Planning
When Women’s Empowerment is the Simple Average of the Indicators 160
Table-5.2.2 Marginal Probability of the Decision regarding Family Planning
When Women’s Empowerment is the Simple Average of the Indicators 161
Table-5.2.3 Results of the Probit Model for the Decision regarding Family Planning
When Women’s Empowerment is Composite Index of the Indicators 165
Table-5.2.4 Marginal Probability of the Decision regarding Family Planning
When Women’s Empowerment is measured by Composite Index 166
Table-5.3.1 Results of the Logit Model for the Incidence of Domestic Violence
When Women’s Empowerment is Simple Average of the Indicators 169
Table-5.3.2 Marginal Probability for the Incidence of Domestic Violence
When Women’s Empowerment is Simple Average of the Indicators 170
Table-5.3.3 Results of the Logit Model for the Incidence of Domestic Violence
When Women’s Empowerment is Composite Index of the Indicators 174
Table-5.3.4 Marginal Probability of the Incidence of Domestic Violence
When Women’s Empowerment is Composite Index of the Indicators 175
xii
Table-5.4.1 Results of the Log-Linear Model for Child Education
When Women’s Empowerment is Simple Average of the Indicators 178
Table-5.4.2 Results of the Log-Linear Model for Child Education
When Women’s Empowerment is Composite Index of the Indicators 179
Table-5.5.1 Estimates of degree of Women’s Empowerment at Household Level 187
Table-5.5.2 Estimates of Composite Women’s Empowerment Index at
Household Level 189
Table-5.5.3 Estimates of degree of Women’s Empowerment at Community Level 193
Table-5.5.4 Estimates of Composite Women’s Empowerment Index at
Community Level 194
Figures
Figure1.2.1 Dynamics of Women’s Empowerment 9
Figure-4.4.1 Percentage Distribution of Women’s Empowerment at
Household Level 146
Figure-4.4.2 Percentage Distribution of Women’s Empowerment at
Community Level 146
Figure-5.3.1 Age-Group Wise Prevalence of Domestic Violence against Women in
Bankura District 167
1
Chapter One _____________________________________________
Women’s empoWerment
An Instrument for Human Development
1.1. Introduction
In the present era, the concept of development has been broadened. In addition to its
quantitative aspects, like increase in real income, consumption, wealth etc. and their
equal distribution, it includes some other qualitative aspects of human livelihood like
capabilities of person, reduction of vulnerabilities, freedom to choose alternative
opportunities etc. Professor A.K. Sen (1999) has explained development in terms of the
expansion of real freedom that the citizens enjoy to pursue the objectives they have and
in this sense, economic development is the expansion of human capabilities. This is
known as capability approach of development. Capability is one kind of freedom, the
range of options that a person has in deciding what kind of life she / he wants to lead. So
a person should have capacity to choose best life that she/he likes from the constraint
opportunities available to her/him. Sen (1999) has also emphasized on the process of
expanding freedom equally for all people. In this view, the meaningful development
includes gender equality in enjoying freedom in addition to less poverty or better access
to basic amenities and opportunities. Following this view in the recent years the
capability approach of development is widely accepted in most of the developing
countries for accelerating the developmental process.
In September 2000, in the Millennium Summit held under the auspices of UNDP, the
member countries promised to fulfill a series of time-bound targets related to human
development with a deadline of 2015. These are known as the Millennium Development
Goals (MDGs). The MDGs have been regarded as the global agreement to combat
2
human deprivation through collective and multilateral actions. The agenda of the MDGs
have been noted below.
Millennium Development Goals (MDGs)
1) Alleviation of extreme poverty and hunger
2) Achievement of universal primary education
3) Eradication of gender inequality and improvement of women’s empowerment
4) Reduction of child mortality
5) Improvement of maternal health
6) Prevention of severe diseases (AIDS, malaria, tuberculosis etc.)
7) Ensuring environmental sustainability
8) Development of global partnership
In Millennium Development Goals (MDGs) summit, 2010, the international
development community recognized gender equality and women’s empowerment as the
major development goal in their own right (MDG 3 and 5) and as critical channels for
achieving the other MDGs and reducing income and non-income poverty. Being a
signatory to the Millennium declarations India has paid due emphasis on inclusive
development in the recent five year plans to articulate policies with view to achieving
MDGs. Several public policies and initiatives in our country reflect that our country is
also committed to develop capabilities equally among human beings.
No doubt the process of capability development approach has changed many things
towards human development during the last one and half decade across the globe. Our
country is also approaching this goal. Despite having some recognised instrumental
power of women’s empowerment, average women in our country like in most of the
countries still have status far behind that of men, particularly in terms of right,
opportunity and endowment. Women constitute almost half of world population. In India
Women constitute 48.46% of the total population (Census, 2011). Like men, women take
the responsibility of human capital formation in the family and hence in building nation
and its destiny; but the status of women in our common society is far below the expected
level. There is a sharp distinction between male and female in our society in terms of sex
ratio, child infanticide, literacy rates, health and nutrition indicators, wage differentials,
3
access to finance, ownership of land and property and in power and culture. Comparing
to their male counterpart, women have lower access for resource, education and health
facility, medical care and lower percentage in earned income, lower power, and lower
percentage even in food. Not only that they are dependent of others, like father, husband
or sons. A common woman does almost all types of works e.g., cooking, tailoring,
sweeping, washing, nurturing, taking care of family members, housekeeping, working as
laundry person, nursing etc in the home and sometimes they do drudgery. She gets very
little time or no time at all to think about herself and her likings which adversely affect
her health and mind also. But the tragedy is that these works do not get any recognition
in our society. Instead of getting the respect they deserve for their roles in the society,
they become vulnerable, marginalised and oppressed population in the society. This
gender inequality can be vanished by enhancing opportunities and capabilities of women
i.e., by promoting empowerment among women. Against this end enhancing women’s
empowerment is recognised as a social movement. Simply women’s empowerment is the
generation of decision-making power among women and providing opportunities to
them so that women can precisely decide about themselves and about the world
surrounding them.
The convention in our patriarchal society is that male persons enjoy the freedom of
taking decisions about all types of household matter and social matter and women only
obey the decisions. Although this convention has been changing recurrently, but it does
not happen at that rate as we need. We shouldn’t forget that nearly half of world
population is women and a major proportion of poor is women. Women are poor not for
the lack of their ability to participate in the production process but for the lack of
opportunity to participate in the production process or for non-recognition of their
housekeeping activities as productive. Globally the gender gap in economic activity is
very prominent. Women are, generally, unemployed or underemployed and when
employed, in most of the cases, they are under paid and employed in informal sectors.
Though working hour of women is very high within house, it is not counted as most of
the time women are involved in household’s jobs. These household’s jobs are not
included in economic activity of a country.
4
Different indicators of human development show that women have lesser access to
property, resource, education, health facilities, medical care and lower percentage in
earned income and finally lower participation in job market, if not least, in decision
making power also, (World Bank, 2001). Moreover, they are the victims of domestic
violence. The basic cause behind this poor condition of women is the bypassing of
women population by the most mainstream development activities. A nation with
slightly less than half of women population can never step on the pave of development
by bypassing its women population. Without developing women population,
developmental procedure will become farce. Swami Vivekananda realised this more than
hundred years ago. He said that “There is no chance for the welfare of the world unless
the condition of women is improved. It is not possible for a bird to fly on one wing”. We,
therefore, should not think of true and total development leaving women aside. Further,
women become marginalised and vulnerable section of the society as social customs and
political system deprived them from their right under the guise of cultural and religious
traditions. Women are, also, victimized by the ethnic tradition. The solution to these
problems requires the building of the capabilities among women. For the betterment of
the condition of women it is necessary to empower women. This urgently, needs the
creation of an environment in which the distribution of power and resources, the
opportunity to engage in productive work, opportunities to access education, medical
care and public services can move in favour of women population.
The remaining part of this chapter has been divided as follows. In section 1.2, we have
presented the theoretical idea of women’s empowerment. Section 1.3 shows the existing
status of women. Relevancy of the study has been discussed in section 1.4. Section 1.5
deals the justification of the selection of the district of Bankura for studying the
empowerment of women. In section 1.6 along with its sub-sections we have presented
the socio-economic and demographic characteristics of the population, particularly
women, in the district of Bankura. The objectives of our research work have been
specified in section 1.7. Finally, section 1.8 concludes this chapter.
1.2. Idea of Women’s Empowerment
The term ‘Women’s Empowerment’ becomes a catchphrase in most of the
developmental studies during the era of globalization. The World Bank and many other
5
development agencies have emphasised the concept of empowerment, specifically,
women’s empowerment, in theoretical discussions and policy perspectives. Exercise of
the concept of empowerment has a long history. During the mid-seventeenth century, it
was used as a legalistic meaning ‘to invest with authority’. Thereafter it began to appear
as a more general word with meaning "to enable or permit". The present generation
meaning of empowerment has been coined in the civil rights movement, which sought
‘political empowerment’ for its followers (Tripathi, 2011). The concept of women’s
empowerment has been broadly used in feminist movement across the globe. But we did
not get any particular definition of women’s empowerment from the feminist form of
advocacy. Advocates of feminist movement think women’s empowerment as a synonym of
gender equality and it is an end in itself. They viewed the empowerment of women as a set
of rights towards equality with men (Kabeer, 2001). According to feminist approach
empowerment is entirely a subjective matter. The idea of empowerment as an instrument
of human development came into discussion after 1980s. The new paradigm concept of
women’s empowerment was closely related with feminist discourse. In addition to the
feminist goals the new idea of women’s empowerment had some important policy pay-
offs (Alsop, et al., 2005). It was known as instrumentalist form of advocacy towards
women’s empowerment. Advocates of instrumentalists tried to define women’s
empowerment with a broader vision compared to the vision of feminists. They had tried
to quantify the concept of empowerment. However, this concept has still not been clearly
defined and segregated from other closely related concepts. Sometimes empowerment is
a process of improving the qualitative aspects of women; sometimes it is a state or the
expected effect of empowering process. Different studies have looked into the concept of
empowerment from different angles according to the need of their studies and different
social contexts.
Many studies have used the term empowerment as the process of empowering groups or
individuals (Rowlands, 1995; Molhotra, et al., 2002, Samanta, 2009, Verma 2009). In
our study we concentrate on women’s empowerment at the individual level and at the
community level. Like many other countries, women’s empowerment is considered as a
social movement in India. First of all, we would, vividly, analyze what the empowerment
is. There is an interpersonal variation in meaning and connotation of empowerment
depending on the economic, socio-cultural and on many other conditions of the society.
6
The meaning and the process of empowerment, particularly, women’s empowerment
varies from time to time, region to region, culture to culture. Let us we present some
definition and ideas of empowerment, given by various previous studies.
Literally, empowerment means giving authority or power to powerless in a particular
situation. In UNDP Human Development Report, 1995, women’s empowerment has
been defined as the expansion of choices for women and an increase in the women’s
ability to exercise choices. Empowerment is the power of decision making i.e. autonomy
(Jejeebhoy, 1995). In a working paper of World Bank, Alsop, et al. (2005) state that
empowerment is the enhancing an individual’s or group’s capacity to make choices and
transform those choices into desired actions and outcomes. It is the ability of some
people to control their own destinies even when, their interests are opposed by others
with whom they interact. Women’s Empowerment is a process whereby women become
able to organize themselves to increase their self reliance, to assert their independent
rights to make choices and to control resources which will assist in challenging and
eliminating their own subordination (Keller and Mbwewe, 1991 cited in Rowlands,
1995). Singha Roy (1995) has considered empowerment as a process of creation of
social environment where women can take decision and make choices of their own,
either individually or collectively, for social betterment. Mehra (1997) has defined
empowerment as a process that enables girls and women to challenge current norms and
change conditions. Panda, et al. (2003) have argued that self-perception is the reflection
of empowerment. The process of empowerment refers to power that controls one’s own
life. The study of Khan, et al. (2006) has also explained empowerment as a process for
establishing control over resources and for acquiring ability and opportunity to decision
making process and its implementation.
In the study of Kabeer (2001) we have found a useful definition of empowerment, which
can be applied across the range of contexts. She states that empowerment means “the
expansion in people’s ability to make strategic life choices in a context where this ability
was previously denied to them”. Reddy (2002) defines empowerment as directing one’s
life in such a way that she/he can reach the stage where she/he is more likely to be
successful in whatever she/he attempts to do. According to him, the processes involved
in empowering are increase in people’s awareness and confidence, ability to articulate
7
problems, gaining access to resources and public facilities and negotiating over relations
between different social groups.
According to Lillykutty (2003), empowerment of women develops them as more aware
individuals, who are politically active, economically productive and independent and are
able to make intelligent decision in matters that affect them and their nations. For
empowering women she suggests five hierarchical steps of equality, namely, welfare
(fulfillment of basic needs of women), access (control over power resources such as
personal wealth, land, skill, education, social status, leadership etc.), conscienisation
(eliminating all forms of discrimination against women i.e., taking actions to fill up
gender gap), participation (organizing themselves for being a direct partner of
developmental works.) and control (creating an environment where they can enjoy all
human rights which, are enjoyed by men and ability of women to take decision over their
life and their children). Sen (1998) has argued that poverty led to denial of rights and
opportunities to poor for full participation in society and to arrest this kind of social
exclusion, there is need for improving their capabilities and entitlements, which is
nothing but empowerment.
From the definitions mentioned above the concept of women’s empowerment may be
summed up as follows. The empowerment of women leads to
the generation of decision making power of women in economic, political
and societal issues
increasing access to education; health care; and other public services within
a geographical and social context
eliminating all forms of discrimination against women
creating the sense of dignity among women.
Therefore, women’s empowerment is a dynamic process and is a multidimensional
concept. It is context specific. There are many indicators of women’s empowerment
which may vary from context to context.
8
Quantification of empowerment is relatively a new phenomenon in literature of women
studies. We have seen that different studies have measured women’s empowerment in
different ways. Some studies (Sen, 1999, Molhotra et al. 2002, Handy, 2004, Sridevi,
2005, Adhikary and Dutta, 2011,) have attempted to measure the degree of
empowerment considering several dimensions of women’s empowerment in several
fields. It means that these studies have measured the width of women’s empowerment.
On the other hand a few studies (Kishor, 1997, Becker, 1997, Kabeer, 2001, Alsop, et al,
2005) have proposed to measure empowerment considering the dimensions which
incorporate process of empowerment as well as its effectiveness. Actually these studies
have tried to measure the intensity of empowerment level. According to instrumentalists,
(Sen, 1999, Kabeer, 2001, Jejeebhoy, 2002, Malhotra, 2002, Alsop, et al, 2005) women’s
empowerment primarily depends on two interrelated sets of factors, (1) Agency and (2)
Opportunity Structure. Agency is the person’s ability to make meaningful choices. In
other words, agency is the attribute of the person by which she visualises the alternative
possibilities of action and makes choices. Opportunity structure includes the existing
formal and informal framework within which agents operate. Working simultaneously,
these two factors generate different degrees of empowerment. The degree of
empowerment definitely makes some personal and social welfare which in turn enhances
empowerment. Alsop, et al. (2005) have recommended that the degree of women’s
empowerment can be quantified by assessing the following issues.
(a) Whether a person has the opportunity to make a choice.
(b) Whether a person actually uses the opportunity to choose.
(c) Once the choice is made, whether it brings the desired outcomes.
There is a reciprocal relation between agency & opportunity structure and degree of
empowerment. Further, women’s empowerment has some impacts on development
outcomes. We have represented the dynamic relations in chain diagram 1.2.1 as
developed by Alsop, et al., 2005. It shows that the extent of the agency and opportunity
structure are both the causes and consequences of the degree of empowerment. The
empowerment is an effective instrument of household and social welfare. The
identification of the linkage between women’s empowerment and welfare or
9
development outcomes is the main contribution of the instrumentalists’ advocacy of
women’s empowerment. They have recognized women as the agent of change.
Figure1.2.1 Dynamics of Women’s Empowerment
Source: Alsop, et al. (2005)
However, this explained association, is a theoretical one. There is not any systematic
statistical analysis which confirms this theoretical illustration. Alsop, et al. (2005) have
rightly pointed out that due to paucity of suitable data on direct indicators of
empowerment, the relationship between empowerment and development outcomes still
remains a hypothesis. One of the central objectives of this dissertation is to test this
hypothesis empirically.
1.3. Women in India and in the District under Study: A Fact Sheet
All over the world women remain beyond the reach of development project. In all
components of human development, women’s position is lower than men’s. The census
reports reveal that number of females per thousand males in India has increased from
933 in 2001 to 940 in 2011. Therefore, female population has grown at a rate faster than
that of male population during 2001-2011. It is no doubt a good indicator of gender
equality. This ratio for child under six years is, however, 914 as per census report, 2011
which was 927 in 2001.
Agency
Opportunity
Structure
Degree of
Empowerment
Development
Outcomes
10
Table- 1.3.1 Sex Ratio: Number of Female per Thousand Male
Census 2001 2011
Area Total Rural Urban Total Rural Urban
India 933 946 900 940 947 926
West Bengal 934 950 893 947 950 939
Bankura District 952 952 951 954 954 958
Source: Census Report, 2001 and 2011
In the state of West Bengal sex-ratio is 947 which is record figure of sex ratio in this
state during the last century. However, child sex ratio in West Bengal has reduced from
960 to 950 during 2001-2011. The picture is more or less same in the district of Bankura,
West Bengal. According to census report, 2001over all sex ratio in the district of
Bankura was 952 which increased to 954 in 2011. But it is interesting to note that in the
district, the sex ratios for general castes, scheduled castes and scheduled tribes were 937,
966 and 984 respectively. The figures indicate that general caste community has more
gender discrimination compared to that for the others communities and the scheduled
tribes have positive attitudes towards gender equality. Therefore, in terms of number,
females are closely equal with their male counterpart. The recent trend of sex ratio is no
doubt a good sign; but many other statistical figures frustrate us.
Illiteracy is a social curse on human beings. It adds impetus in the feminization of
poverty and in deterioration in the status of women. India is not free from the curse of
illiteracy. In 2001, the adult women literacy rate of India was only 42.2% whereas the
adult male literacy rate was 67%. In term of enrollments in the educational institutions
they are far below their male counterpart. The combined primary, secondary and tertiary
enrollment ratio among female population in 2001 was only 49% whereas for male
population it was 62%. The literacy gender parity index in India in 1995-1996 was only
0.8. At present 82% of male population aged seven year and above are literate, while
65% of the female are literate (Census, 2011). It is evident that male female gap in
literacy has been declining during the era of globalization. In 2011, 71% of female 7+
aged are literate in West Bengal whereas 60% of female 7+ aged are literate in the
district of Bankura. Women’s education level is lower compared to that of men
population in the state of West Bengal and in Bankura district. It is reported that 60%
women are literate whereas 81% men population are literate in Bankura district (Census
2011). The access to education among women in the area under study is not satisfactory
11
compared to men. In this district 937 girls per thousand boys enrol in the primary
section. Moreover, dropout among girls is higher than that among boys within the
primary section.
Violence against women is a serious social problem across the countries. Violence
against women is now daily news in our country and in our state. Women face violence
not only outside home, but also within home. In accordance with the report of National
Crime Record Bureau (2013) 106527 women in 2012 have reported the experience of
domestic violence from husband and/or in-laws. This number has increased by 7.5%
compared to the figure in 2011. In the state of West Bengal, domestic violence has been
reported to increase by 18% during 2011-12. In terms of this rising rate West Bengal
stands first. The number of dowry death has not been included in this figure. Normally
women in our society do not want to confess the experience of domestic violence. Only a
small section has reported their experience of domestic violence. Moreover, a number of
females are suffering from violence in their natal house. Therefore, the actual prevalence
of domestic violence against women is very high in India.
The survey report of UNIFEM (2011) tells us that one fourth of the Indian women faced
physical violence at home in 2010 and 37.2% of sample women reported that they have
faced domestic violence at least once in their life. In most of the cases violence came
from husband or from close relatives. National Health and Family Survey 2005-06 has
revealed that 26% of Indian women have faced the incident of physical violence in home
within two years of their married life.
West Bengal has got some defamation in this respect. According to National Crime
Record Bureau (2011) West Bengal ranks first in term of percentage of share of crime
committed against women in India. In this state the rate of rape, kidnapping and
abduction, dowry death, cruelty by husband and relatives are 2.6%, 4.1%, 0.6% and
21.6% respectively. Based on 2008 statistics of the National Crime Records Bureau,
West Bengal has the highest number of battered wives among the Indian states. In
dowry-related deaths, it comes fifth among the states. In the district of Bankura reported
number violence against women in 2001 was 210 that increased to 326 in 2005.
12
So far, in India the protection of women from domestic violence Act has been passed in
2005 which has come into force on October, 2006. But due to improper utilization and
implementation of this act and due to our social immoral value a large section of women
in India and in our study state and district are suffering from domestic violence. There
are several forms of domestic violence. It scratches the root of self-confidence and self-
esteem of the woman. Reduction of self-confidence and self-esteem increases the
dependency of the woman on husband or father which is main pillar of patriarchal and
feudalistic nature of our society. Therefore, the root cause of domestic violence lies in
the framework of the patriarchal and feudalistic nature of the society.
However, women are not dependent of men in real sense. Most of them work hard for
their family particularly housekeeping. Household jobs are treated as household duties
not as economic activities. According to the report of NSSO (2009-10) 347 women per
thousand in rural area and 465 women per thousand in urban area are completely
engaged in household duties. On the other hand, only 5 men per thousand in rural area
and 4 men per thousand in urban are engaged in household duties. The majority of the
homemaker women earn some money from outsources working at home. But it is
omitted from the mainstream labour force because these women offer these earning to
their family heads or husbands and earning is looked as family income. It is a fact that
women work more time than men but they earned a very low percentage of income. The
female work time as a percentage of male work time for the year 2000 was 117. But
male population spends 61% of total work time on marketed activities whereas women
population spends only 35% of their work time. Actually women spend most of their
work time in unpaid household works. Their activities are not accounted in the National
Income statistics. As a result women has lower share in earned income. In India the
estimated earned income by women population was only $1531 in 2001 whereas it was
$4070 by Indian male. Economic dependence of women resists them to become a part of
the building of nation.
Further, in India, women are far below their male counterpart in term of access to and
control over resources, participation in work force and remuneration. The discrimination
of the dignity of work of men and women has been focused in their wage differentials.
The daily wage of women labour is Rs 119.76 in rural area, which is Rs 297.35 for men
13
labour (NSSO, 2009-10). Although the participation of women in unorganised sector is
noticeable, their participation in organized service sector is negligible. Census report
2001 reveals that in Bankura district 32.04% women participates in the work force.
However, women’s participation in the manufacturing and service sector in this district
is not commendable. Women of the district are mainly engaged in the primary sector.
But their share in landholding is insignificant (Bankura District HDR 2007).
In 2001, 39.44% of total seats in Panchayet Bodies were reserved for women. Though
women physically filled up the seats, in most of the cases the male leader influences
their activities. In our country, state and sample district the incidence of child marriage is
rampant. In West Bengal 53% of women marry before reaching the legal minimum age
at marriage (NFHS-3). In our country average age of marriage is 22.2 years while in
West Bengal it is 20.3 years. It is evident that average age of marriage for women in
Bankura lies in 15-18 years. Bankura District HDR, 2007 has reported that there is social
discrimination among women either on the issue of ‘caste’ and untouchability or on the
basis of ‘haves and have not’s in rural Bankura.
Further, the access to health care facilities by women is not satisfactory in India. The
pregnant women receiving prenatal care in India was 62% in 1996 while it was 70% for
the whole world. It is reported that 63.2% of women are suffering from anemia. NFHS-3
has reported that 22% of men think that contraception is women’s business and a man
should not have to worry about it. It has been reported that 37% of women have used
sterilization as a method of family planning while it is only 1.2% of men.
In the light of the above information we can conclude that number of women is
increasing but the status of women in our society is far below the expected level. A large
part of them could not enjoy the minimum facilities which are necessary for their health
and dignity. Women in contrast to men have low access to property, community
resources and services, education, health facilities, medical care. They have lower
percentage in earned income. They are vulnerable and marginalized group compared to
men in society. A large number of women face domestic violence. Therefore, it is
essential to abolish the gender inequality against women for inclusive development of a
country.
14
1.4. Relevance of the Study
Observing the present status of the women in India and in the district of Bankura we find
that globalization fails to develop the status of women population to the expected
standard. Women population, particularly in India, has been excluded from the
favourable impact of globalization. In the Millennium declaration, 2000, the member
states of United Nations declare that for true sustainable development it is necessary to
abolish poverty, hunger, disease etc. and announce the empowerment of women as one
of the Millennium Development Goals. Literally, empowerment means to give some
power in the hands of powerless. It enriches person with power. Many developmental
projects have been undertaken by the governments of various countries. But the fact is
that most of these mainstream developmental projects, latently or actively, bypass
women population. Even when mainstream development projects include women by
increasing investment on women’s health and education, it is for lowering birth rate and
to improve the well being of children. These are the policies not for women’s own well
being so that they can expand their ability to exercise choices (Mehra, 1997). As a result,
women population remain far below the male population not only in terms of numbers,
but also in terms of various development indices such as literacy rate, employment,
access to medical facilities, enrollment in educational institution etc.
Time has, now, come to take special policies, exclusively, for the development of
women. Enhanced women’s empowerment can serve as one of the strongest weapons to
fight against such obnoxious reality. Empowerment is now considered as a component of
human development. United Nations declared the year, 2001, as the year of women’s
empowerment. Women can achieve better familial, societal and economic status and can
fight against the various atrocities if some policies are taken to empower them. This
requires the creation of an environment in which the distribution of power and resources,
the opportunity to engage in productive work, opportunity of access to education,
employment, medical care and health services etc. can move in favour of women
population. To quest for the suitable policies for empowering women it is necessary to
find out the responsible factors of women empowerment and their importance to
improve empowerment. Increased level of empowerment among women increases the
importance of women in their family and in the society. This may be the panacea of the
two dangerous diseases in our society namely the gender inequality and poverty.
15
Therefore, understanding the concept of empowerment and the study of the determinants
of women’s empowerment are very much relevant in present context.
In the existing literature still now women’s empowerment is a fuzzy concept. In recent
times, the instrumentalist form of advocacy has been translating the feminist views into
the mechanists’ discourse of policy (Kabeer, 2001). There have no universally accepted
indicators of women’s empowerment in literatures. We find different set of indicators of
women’s empowerment in different studies depending on the context of the study. These
may be helpful to construct the empowerment index. Yet a major portion of the existing
studies are anecdotal, informative and descriptive. They have tried to present a
theoretical explanation of the relation among intermediary indicators of empowerment,
agency and opportunity structure and development outcomes. Several studies have
quantified women’s empowerment in several contexts. But the study of the impact of
women’s empowerment on development outcomes is not very much common in the
existing literature. In a working paper of World Bank Alsop, et al. (2005) has rightly
mentioned that
“while we currently have much anecdotal and case study evidence to
suggest an instrumental purpose in empowering people, robust data
demonstrating a clear association between empowerment and development
outcomes are hard to find”.
This argument justifies the concrete relevance of our study which deals with the
estimation of the impact of women’s empowerment on development outcomes.
Quantification of the concept of women’s empowerment is the main contribution of the
instrumentalists. But there is not any unanimously accepted measure of women’s
empowerment at the individual level. It is not a surprising fact that the nature of
individual’s empowerment absolutely depends on existing infrastructure that the
individual may get access. In other words, the nature of women’s empowerment is
context and area specific. It indicates the difficulty of the proposition of universal
measure of empowerment at the individual level. So, for an area specific study one
should be conscious regarding the measure of empowerment. This fact raises the
16
relevance of our study. In our dissertation we have attempted to develop a measure of
individual empowerment for the women in the district of Bankura, West Bengal. We
have planned to measure the women’s empowerment at the household level and at the
community level.
The instrumental advocacy of women’s empowerment argued that women’s
empowerment have some positive impact on household and child welfare. But the
impact study of empowerment on development outcomes is not common in the existing
literature. World Bank (2005) has explained the theoretical relation between
empowerment and development indicators. Theory at best can give the nature and
direction of the relation among the factors. However, without an empirical estimation of
the relationship, theory can be of little use in policy. Therefore, empirical study relating
to the impact of women’s empowerment on household and child welfare is immense
important in India. It makes our study a relevant one.
The importance of family planning has got attention in our country since the early 1950s.
But how the attitude of rural couples towards family planning is related with
empowerment of women is not well established. In recent times the problem of domestic
violence against women has also received deep attention of the governments. It is
expected that women’s empowerment is an important determinant of the incidence of
domestic violence. Therefore, in the present era, examinations of the impact of
empowerment on family planning decision and on incidence of domestic violence
against women are important issues which are the central objectives of our dissertation.
Recently, in addition to women’s welfare the issue of child welfare has come forward as
an aspect of inclusive development. Towards this end, our study regarding the impact of
women’s empowerment on child education is most relevant. Hence, in order to
understand the level of empowerment of the rural women level and its impacts on
household and family welfare, our study is no doubt important.
1.5. Justifications of the District Selection
In order to study the nature of empowerment of women and its impact on household
well-being we have considered the district of Bankura in West Bengal. The justifications
behind the selection of the district are stated as follows.
17
Bankura district is one of the poor and backward districts in India. In this district all the
components of human development index is lower than the all India average. It means
that people of this district are lagging in position relative to average Indians and average
people of West Bengal. The condition of women is worse than that of men. In section-
1.3 we have presented the required information for showing that the women of Bankura
district are in the lower position than men in all dimensions of human development. To
get a vivid picture of the position of women in this district we need to study about their
empowerment at the household level and at the community level.
From many primary and secondary sources we know that more than half of the
households in the district of Bankura belong to lower social castes. Conventionally
women of these lower social castes work outside home for earning their livelihood.
However, most of them participate in unorganized sector like agriculture and animal
husbandry. So far, our primary observations indicate that although these women support
their respective household economically, a large section of them suffers from domestic
violence. It is observed that women get less importance in taking decisions on various
issues of family welfare, like family planning decision and spending for their children’s
education. These observations are odd enough and motivate us to study the attitudes of
women towards family planning, incidence of domestic violence against women and
spending towards education for households in the district of Bankura.
Moreover, in spite of the economic contribution of these women a part of them fails to
enjoy the decision making power in different economic and non-economic fields within
and outside household. During the course of this study it would be possible to find out
the causes of the powerlessness of women and the different factors responsible for this
powerlessness. These help the policy makers, the government, the non government
organizations take appropriate projects and policies for improving the situation of
women in our study area. All these justify the selection of the district under study.
1.6. Socio-Economic-Demographic Characteristics of Bankura District
In order to study the issues regarding empowerment of the rural women we have selected
the district of Bankura in West Bengal. With this end in view, first we need to
understand the socio-economic-demographic profile of the district, which are relevant to
18
carry out the research work. These statistics will provide us a clear picture about the
relevance of the study and consideration of the district and thereby induce us to
empirically estimate the impact of women’s empowerment and thereby determining
factors of it. In the following sub-sections we would introduce the district under study.
1.6.1. Geographic and Administrative Profile
Geographically, the district of Bankura is fourth largest district in West Bengal. In terms
population as per census 2011, it ranks thirteenth. It is located in the western zone of
West Bengal, which is called ‘Rarhanchal’. A major part of ‘Rarhanchal’ is known as
‘Jangalmahal’. The district of our study belongs to the zone of ‘Jangalmahal’. Once
upon a time, a part of the district was kingdom of the Malla Raj. Present administrative
area of Bankura District took its shape on 1881 under the Burdwan Division of West
Bengal. The geographical area of the district is 6882 square kilometres out of which
6820.51 square kilometres are rural. The geographical statistics of Bankura district have
been depicted in table-1.6.1.
Table-1.6.1 Geographical Area (Square Kilometres)
West Bengal Bankura District
Rural 85427.26 6820.51
Urban 3324.74 61.49
Total 88752 6882 Source: Director of Census Operation, West Bengal, 2005.
This district is bounded by Burdwan district in the north, Hooghly district and Burdwan
in the east. The west line of Bankura district is bordered by the district of Purulia. South
line of this district is a border line with the district of Paschim Medinipur. Major part of
this district particularly western part is mainly undulating terrain. ‘Susunia’ and
‘Biharinath’ are remarkable hill of this district. Land of the eastern part is relatively plain
and fertile. In the district of Bankura major portion of the land area is structured by
laterite soil and light forestry, which is not suitable for intensive agriculture. Basically,
the district is draught prone, only Damodar and Kangsabati river projects provide
irrigation to a very small part of the land area of the district. Mundeswari, Dwarakeswar
and Damodar rivers also create some fertile valley in the district. Dwarakeswar river
flows through the middle of the district and Damodar river separates the district from
19
Burdwan district. Yet, 82% of total population has engaged themselves in cultivation
because there is no better opportunity in the other sectors. Not only that, in the course of
our pilot survey we have experienced that the means of transportation and condition of
the roads in this district are too bad and remote. These prime features no doubt influence
the socio-economic-cultural status of the people of the district.
Table-1.6.2 Administrative Profile of Bankura District
Sub-division 3 (Bankura Sadar, Bishnupur, Khatra)
Community Development Block 22
Police Station 22
Municipality 3 (Bankura, Bishnupur, Sonamukhi)
Gram-Panchayat 190
Village 5178
Inhabited Mouza 3543
Uninhabited Mouza 385 Source: Human Development Report, Bankura District, 2007
We have depicted the administrative system of the district of Bankura at a glance in
table-1.6.2. Administrative headquarter of the district is situated at Bankura town.
Bankura district is divided into three sub-divisions namely – Bankura, Khatra and
Bishnupur. Bishnupur was the capital of the kingdom of the Mallaraj and it has a cultural
heritage. There are twenty-two Community Development Blocks and twenty-two Police
Stations in the district. The three-tier panchayet system of the district constitutes of 190
Gram Panchayets with 03 Municipalities, 22 Panchayet Samities and one Zilla Parishad.
There are 5187 villages and 05 towns in this district.
1.6.2. Demographic Characteristics
In 2011, total population of the district of Bankura stood at 35, 96,292, which is 3.93%
of the total population of West Bengal. Out of them 91.64% live in rural area (Census
Report, 2011). This figure confirms that most of the people of this district remain far
away from urban amenities. Census Report, 2011 shows that 1608635 females live in
rural Bankura, whereas 147153 females live in urban Bankura. Based on the census
report, 2011, we find that in the district of Bankura the decadal growth rate of population
in between year 2001 and 2011 has been 12.64 percent, which is slower than that of our
state and nation. Perhaps, it may happen due to the continuous emigration of the people
20
from the district. In 2001, rural population in Bankura district was 92.63 percent of total
population, which has marginally decreased to 91.64 percent in 2011. The figure
establishes that almost all people of the district are residents of rural areas. It is
interesting to note that decadal growth of urban population in Bankura population
increased to 8.36% in 2011 from 7.37% in 2001. It means that the rapid urbanization
take place during the last decade. The rural urban population distribution is shown in
tables 1.6.3 and in 1.6.4.
Table-1.6.3 Population and Decadal Growth rate by Residence
Total Rural Urban
Person 3596292 3265613 330679
Male 1840504 1686978 153526
Female 1755788 1608635 147153
Decadal Growth rate (person) (per cent) 12.64 11.43 27.8
Source: Provisional Census Report, 2011
Table-1.6.4 Percentage Distribution of Rural Urban Population
State/District 2001 2011
Rural Urban Rural Urban
West Bengal 72.03 27.97 68.11 31.89
Bankura 92.63 7.37 91.64 8.36
Source: Census 2011, Provisional population totals, Director of Census Operation, W.B, 2005
Table-1.6.4 shows that in between 2001 and 2011 percentage of rural population has
declined in West Bengal as well as in the district of Bankura. However, all the urban
areas in the district are dependent on agriculture and its allied activities. One distinct
feature of the district population is that a good number of rural families are seasonal
migrant and move as agricultural labourer to the neighboring districts at the time of
cultivation (Our field survey).
Table-1.6.5 reveals that sex ratio in the district of Bankura is better than that in the state
as a whole. Sex ratio in the district has slightly improved from 952 females per thousand
males in 2001 to 954 females per thousand males in 2011. In the state as a whole this
improvement is good enough during the last decade. It is a positive sign towards gender
equality. However, the rate in district is much slower than the rate of our state where
sex-ratio increases to 947 in 2011 from 934 in 2001.
21
Table-1.6.5 Sex Ratio in West Bengal and Bankura District
State/District 2001 2011
West Bengal 934 947
Bankura District 952 954
Source: Census Report 2011, Government of India.
Table-1.6.6 Caste Wise Population (Percentage) with Reference to 2001 Census
State/District
Scheduled Caste Scheduled Tribe General & Backward Classes
West Bengal 23.01
5.49 71.5
Bankura District 31.25 10.36 58.39
Source: Director of Census Operation, W.B.2005.
This sex ratio of the district is different for different caste categories as shown in table-
1.6.6. It is 937, 966 and 984 for general, SC and ST population respectively (Census,
2001). This shows that the higher the caste in social status the higher is the gender
discrimination. In the district of Bankura 31.25% of total population belongs to
Scheduled caste category and 10.36% belongs to Scheduled tribe category (Census,
2001). This confirms the rampant presence of backward classes in this district. These
figures are relatively high compared to the figures of our state West Bengal. Thus the
district of Bankura is the residence of a large number of Scheduled Caste and Scheduled
Tribe people of West Bengal.
1.6.3. Socio-Economic Characteristics
According to West Bengal Human Development Report (2004) Bankura district ranks
eleventh in human development among seventeen districts in West Bengal (ignoring the
breakup of Dinajpur and Midnapore districts). The values of human development indices
have been depicted in table 1.6.7.
Table –1.6.7 Human Development Index of Bankura District
State/District Health Index Income Index Educational Index HDI
West Bengal 0.70 0.43 0.69 0.61
Bankura District 0.67 0.26 0.62 0.52
Source: West Bengal Human Development Report, 2004.
22
It is seen that values of health index and educational index for the district under
consideration are slightly lower than that for West Bengal but in terms of income index
our district lies far behind the state. Therefore, income poverty is more serious than the
poverty in terms of health and education in Bankura district relative to that in the state of
West Bengal as a whole. According to NSSO (1999-00) 59.62% of rural households and
52.38% of urban households in Bankura district were below the poverty line, determined
by planning commission (West Bengal HRD, 2004). Per capita monthly consumption
expenditure is an important indicator of the economic condition of people. National
Sample Survey (1999-00) has reported that in Bankura district per capita monthly
consumption expenditure was Rs. 350.28 for rural people and for urban people it was Rs.
500.40. Therefore, the district of Bankura is a poor district in West Bengal.
Table-1.6.8 Percentage of Population According to Occupational Status in 2001
Year Farmers Agricultural Labourers Non-farm Occupation
1991 43.2 36.5 20.3
2001 32.6 37.1 30.3
Source: West Bengal Human Development Report, 2004.
From the table 1.6.8 it is clear that most of the people in Bankura are engaged in farm
activity. It is an indicator of underdevelopment. The table indicates that during the period
1991-2001 a good percentage of population has shifted themselves from farm to off-farm
occupation. It is a good indication for development. So the district under study is a
developing district of West Bengal. However, the percentage of agricultural labourers
remained more or less same during the period 1991-2001. This means that the
agricultural labourers could not change their occupational status. Only some farmers are
shifted to off farm occupation. Therefore, agriculture is the main occupation of the
people of this district. We find that 32.6% of total population works as farmer, 37.1% as
agricultural labourer and only 30.3% engage in off-farm jobs. Hence, nearly 70% people
directly depend on agriculture for their livelihood.
1.6.4. Educational Status
Table-1.6.9 has shown the gender-wise education of the people in West Bengal and in
the district of Bankura. Though universal human right declaration treat education as one
23
of the basic right of every individual, the light of education does not reach to a large
proportion of population in Bankura district. On the basis of final census data 2011, the
overall literacy rate in India is 74.4%, where as it is 77.08% in West Bengal and 70.95%
in Bankura district. Female literacy rate in Bankura district is also lower than that at the
state and national level. The table reveals that literacy rate of Bankura district has
increased during the decade 2001-11 by nine percentage point in respect to total
population. Yet, after the twenty years of total illiteracy drive program we have seen that
in the district of Bankura 29 % of total population aged not below six year could not read
and write. According to Bankura Human Development Report, 2007, female literacy rate
is less than 50% in more than 16 out of 22 blocks of this district. This implies that major
proportion of rural women is illiterate. This report reveals that the girls’ enrolment per
thousand boys is 937 in this district in primary and upper primary level and with higher
classes, girls’ enrolment decreases and thereby, dropout increases.
Table 1.6.9 Gender Wise Distribution of Literates
State/district 2001 2011
Total Male Female Total Male Female
India 65.4 75.8 53.67 74.4 82.14 65.46
West Bengal 68.2 77.02
60.2 77.08 82.68 71.16
Bankura District 63.44 76.76 49.43 70.95 81.00 60.44
Source: Census 2011, Provisional population totals, Director of Census Operation, W.B, 2005
The table-1.6.10 indicates that there is inequality in literacy rate among scheduled castes
and scheduled tribes people. According to census data 2001, the literacy rate is 58.22%,
27.11% and 42.92% in respect of male, female and total scheduled castes population
respectively. These percentages are lower than the corresponding percentages for West
Bengal as a whole. The literacy rates are 67.84%, 31.13% and 49.60% in respect of
male, female and total scheduled tribe population respectively in Bankura district.
Table-1.6.10 Sex Wise Literacy Rate among Scheduled Caste Scheduled Tribe
State/District Scheduled Caste Scheduled Tribe
Male Female Person Male Female Person
West Bengal 70.54 46.90 59.04 57.38 29.15 43.40
Bankura District 58.22 27.11 42.92 67.84 31.13 49.60
Source: Director of Census Operation, W.B.2005.
24
Based on the statistics we can say that the scheduled tribe communities of Bankura
district are educationally advanced compared to other part the state. The percentage of
literate persons among tribal people is higher than those among scheduled castes people.
Both male and female people of scheduled tribes are better in formal education than
those in scheduled caste community. But females in both categories have low access to
education in contrast to their male counterpart. This information reveals that the SC and
ST women in this district are disadvantageous group in contrast to the men population.
It leads to lower productivity and lower social and professional status of the women of
Bankura district.
1.6.5. Empowerment Statistics
According to census 2001, the female work participation in India is 25.7 percent against
the male participation rate of 51.9 percent. In Bankura district female work participation
is 32.04 percent, which is higher than national level. Among the female workers, 21.66
percent work as cultivators, 48.74 percent work as agricultural labourers, 9.72 percent
work in household industry and 19.88 percent are engaged in other activities. This
occupational pattern of women of Bankura district indicates that most of the women
work in fields as agricultural labourers and women’s participation in tertiary sector is
very negligible. This is an indicator of rural poor economy. We have collected all these
information from Bankura HRD, 2007. Women’s work participation has been considered
as one of the determinants of women’s empowerment.
Though Indian constitution does not have any provision of gender bias, there is no
significant presence of women in the field of politics. According to UNDP, 1995,
women’s political participation is considered as one of the major components of gender
empowerment measure (GEM). According to the 73rd
amendment Act in 1972 one-third
of total seats in panchayats and municipalities are reserved for women. Bankura HRD,
2007, tells that 39.44 percent seats of panchayet bodies were reserved for women in
Bankura district in 2001 and it is 6.10 percent more than minimum reservation. In most
of the cases women’s political decisions are influenced by their husband, father, son or
other male family members; women are silent spectators in the field of politics. Actually
they are not active participants of politics; women’s presence in politics is symbolic,
ineffective and latent.
25
Bankura district Human Development Report, 2007, shows that women in this district
suffer from protein deficiency and malnutrition and from anemia and tuberculosis. Most
of the women do not get antenatal care. In deed women are deprived of their basic right
of health. HDP, 2007 of Bankura district has reported that the average age of marriage
for women lies below18 years. However, it is above 18 years for general caste women.
Dowry system where bride’s parents pay dowry to groom’s family, is very much
common among the general and scheduled castes marriage. It is reported that the dowry
for general caste ranges from Rs. 50000 to Rs. 400000 in 2005. It is important to note
that tribal community has no such dowry system. In tribal community groom’s family
give some jewellery to the bride as gift. Like any part of the world the women in the
district of Bankura are not free from violence. In 2005, police stations in Bankura district
have reported 326 cases of domestic violence against women.
1.7. Objectives of this Study
Prior to implementation of any development project, it is essential to measure how much
it would be effective to improve the situation of poor, women, lower-caste and
vulnerable and non-empowered population. After complementation of project, it is
equally important to study the impact of the project on the people whom it was
implemented for. So it needs to develop an index of empowerment, which would include
most of the indicators of empowerment. Empowerment is not only qualitative in nature;
it has a number of definitions also depending on the context of the study. Meaning of
empowerment varies from person to person and region to region, caste to caste, ethnicity
to ethnicity. So it is not very easy to propose a common measure for empowerment or
quantify it. We want to divide our whole study into three steps. First, we have tried to
quantify women’s empowerment. Second, we assess the impact of women’s
empowerment on household welfare and third step is the exploration of the determining
factors of women’s empowerment in the district of Bankura, West Bengal.
In the quantification of women’s empowerment we are interested to measure it at the
household level as well as at community level. There are several dimensions of women’s
empowerment at the household level as well as at the community level. We, therefore,
have to select the indicators and dimensions of the women’s empowerment and explain
the logic behind the selection. With the use of the indicators of the women’s
26
empowerment we form an index by which we can measure empowerment, women’s
empowerment in particular at the individual level. We have planned to formulate
empowerment index for both at the household level and at the community level. The
detailed methodology has been explained in chapter three. It would be helpful to
evaluate the impact of women’s empowerment on household and child welfare that has
been considered as a result of empowerment.
In order to assess the impact of women’s empowerment on household and child welfare
in the district of Bankura we have considered three issues – attitudes towards family
planning, incidence of domestic violence and spending for child education. No doubt all
the three aspects of household and child welfares are closely related with women’s
empowerment. But the question that will be naturally asked is that to what extent or
magnitude the issues mentioned above are in reality properly executed or materialised
through women’s empowerment.
Although India has taken several steps towards family planning since 1952, still now
more than 40% women did not use family planning measures. This picture appears better
in West Bengal in contrast to that in major states of India. But still here 29% of women
in reproductive age did not take decision regarding family planning. More surprisingly,
in our country seven percent women do not adopt any short of family planning measure
just in order to have maximum number of children. Fifteen percent of women reported
their unwillingness to adopt family planning measure either due to the disapproval of the
decision making family members or their religion, and another ten percent reported
concerns about health or the fear of side effects (NFHS-3). Some women are reluctant to
adopt family planning measures until they get male child. This information we have got
in our field survey in the process of personal interview. These reasons behind the
unwillingness to adopt using family planning measures are directly linked with the
unawareness, unconsciousness and lack of empowerment on the part of women.
Therefore, women’s empowerment at the household level and at the community level is
expected to affect the decision towards family planning. With this end in view, we have
planned to examine the impact of women’s empowerment at the household level and at
the community level along with other household and community characteristics on
decision regarding family planning for the women in Bankura district, WB.
27
Our second aspect regarding household welfare is domestic violence against women.
Violence against women is common irrespective of social customs, economic status,
caste, creed, religion and other cultural backgrounds. The issue of domestic violence has
been emerged as a research agendum in the field of social science as well as in medical
science. Actually, the manifestation of patriarchal power results in violence against
women. In a patriarchal country like India the male enjoys dominance, privileges and
freedom in all aspects of life, where as women are deprived of basic human rights and
therefore are the victims of social customs. The definition of violence against women
(VAW) cannot be singled out. UN General Assembly, in its resolution 48/104 of 20
December 1993 defined VAW as “Any act of gender based violence that results in or is
likely to result in physical, sexual, psychological harm or suffering to women including
threat of such act, coercion or arbitrary deprivation of liberty whether occurring in public
or private life”. Any untoward action that adversely affects security, freedom and
welfare of women, is considered as violence against women. It gets the women deprived
of enjoying their basic human rights.
The victims of domestic violence range from the killing aged woman to killing of female
foetal. Domestic violence is present everywhere irrespective of time and space, region,
religion, class, caste, status and position, economic condition, age etc. The tragedy
becomes aggravated because this crime does occur not only in external atmosphere but
also within the family at home. When violence occurs within the family, this is
commonly known as Domestic Violence (DV). World Health Organization (WHO) has
defined DV as “the range of sexually, psychologically and physically coercive acts used
against adult and adolescent women by current or former male intimate partners”.
Beating, kicking, slapping, harmful restrictions to regular ordinary behaviour, normal
activities and freedom of movement, threats to demolish from property, denial of access
to resources and control over assets, threat of murder, hurting the feelings of wife by
using the putrid idiom, sexual assault, rape, provocation to commit suicide or deliberate
self-harm etc. are some examples of domestic violence. It affects not only the woman
concerned but also affects the family as well as society.
We have already mentioned that a large section of women suffers from domestic
violence in spite of the existence of the Act for the protection of women from domestic
28
violence. It shows that laws and order are not sufficient to combat the incidence of
domestic violence against women. In order to find an alternative way we would like to
investigate the association between the incidence of domestic violence and women’s
empowerment. We actually like to assess the impact of women’s empowerment at the
household level and at the community level along with other socio-economic factors on
the probability of the incidence of domestic violence against women.
The issue of child welfare is equally important in the goal of inclusive and sustainable
development. In our patriarchal society, no one would deny the role of mother in child
welfare like in child education, health, nutrition and his/her all round development.
Women’s empowerment is, therefore, a crucial factor for child welfare. In order to
examine the role of women’s empowerment in child education we consider the issue of
spending on education of the children of the sample women in the district of Bankura.
Particularly, we are interested to explore the impact of women’s empowerment at the
household level and at the community level along with some selected household
characteristics on spending for her child.
Once we find that empowerment has an instrumental effect on household and child
welfare, we need to search for the determinants of the women’s empowerment and need
to develop a theoretical relation between the women’s empowerment and its
determinants. Once we establish the theoretical relations between women’s
empowerment and its determinants we should test the empirical validity of these
relations. It can also be used for policy making about how much and what kind of
incentive is necessary to improve the empowerment level of women of a particular
region.
In this part of our study we have intended to estimate, empirically, the empowerment
level of women of Bankura District. Using a set of primary data, we want to explore the
significant determinants of the women’s empowerment. In this study, we examine the
effects of several individual and community characteristics like age, education,
occupation, income, financial inclusion, membership in SHGS, caste etc. on women’s
empowerment. This study tells us what kind of other incentives or policies are required
29
in the particular region under study to improve the index of empowerment. We have
designed the study with the particular objectives as follows.
First, we would like to study the possible dimensions and indicators of the women’s
empowerment at the household level as well as at the community level for the
women in the district of Bankura. With reference to the selected dimensions and
indicators we have developed the index of women’s empowerment at the household
level as well as at the community level for each sample woman.
Second, once we have the women’s empowerment as quantitative variable we can
estimate the impact of women’s empowerment on household and child welfare. In
this step we have considered three aspects of household and child welfare – decision
regarding family planning, incidence of domestic violence against women and
spending for child education.
a) We, therefore, investigate the effect of women’s empowerment at the
household and at the community level along with the other socio-
economic and demographic traits on the decision regarding family
planning for the sample women in the district of Bankura.
b) This study assesses the impact of women’s empowerment along with
some selected factors on the incidence of domestic violence against
women in the district of Bankura.
c) We also seek to examine the impact of the women’s empowerment along
with other household and community characteristics on proportion of
household income spending for children’s education for the households
in the district of Bankura.
Third and finally, we intend to trace out and examine the responsible and
significant factors in the determination of the women’s empowerment at the
household level as well as community level. It will help us develop an empirical
relation between the index of empowerment and its determinants.
30
1.8. Conclusion
Discussion of this chapter clears that empowerment of women and abolishing of gender
inequality is essential for inclusive and sustainable economic development. In UNDP
Human Development Report, 1995, the main slogan was “Human Development if not
Engendered is Endangered.” To empower women many steps, policies and projects have
been taken by various international agencies and governments of many countries, but
these were inadequate. Under this backdrop, it is necessary to suggest and implement
appropriate policies for empowering women. What kind of policy should be effective for
empowering women depends on social, economic, demographic characteristics and
cultural norms of the region where the women live. Policies of empowerment are region
specific and culture specific. We would like to investigate the nature, the dimensions and
determinants of women empowerment in the district of Bankura, West Bengal. We have
also planned to investigate the impact of empowerment on household and child welfare
indicating the decision regarding family planning, incidence of domestic violence against
women and proportion of household income spending for child education. In this chapter
we, mainly, chalk out the concept of empowerment, the relevancy and objectives of our
study and social-economic-demographic characteristics of our study area. The detailed
review of the literature for the dissertation has been presented in the second chapter.
Later on, in the third chapter, we present the various dimensions of women’s and a
proposed measurement of the women’s empowerment. The third chapter deals with the
methodology of estimation and hypotheses for our study. We have interpreted and
discussed the empirical findings of this study in chapter four and five. On the basis of
this empirical study, finally, in chapter six we suggest some alternative policies for
improving women’s empowerment and family welfare of the rural households in the
district of Bankura.
31
Chapter two
Review of Literatures on Women’s
Empowerment
2.1. Introduction
Review of the existing literature is the basement of any research. At the initial stage of a
fundamental research, a researcher should go through the relevant literatures. This
chapter presents a review of the existing literature relating to women’s empowerment
which is the central theme of our dissertation. We have planned to develop a measure of
women’s empowerment. So we need to review the existing definitions and measures of
women’s empowerment. We have to know what kinds of difficulties the previous studies
have faced in measuring women’s empowerment; what the different indicators of
women’s empowerment are; what the existing policies are to accelerate women’s
empowerment, how much they are effective to change women’s empowerment etc.
Understanding the impacts of women’s empowerment on different developmental
outcomes is another objective of our study. We need also to understand the possible
determinants of women’s empowerment. In order to gather the existing knowledge
regarding the above mentioned facts we need an extensive literature survey. Keeping this
view in mind, we have reviewed various literatures relating to our study.
The remaining part of this chapter has been structured as follows. In section 2.2 we have
presented the studies, which deals with the conceptual framework of women’s
empowerment. The studies relating to the impact of women’s empowerment on
household welfare have been reviewed in section 2.3. Section 2.4 cities the empirical
studies on women’s empowerment. This section is divided into two sub-sections where
we present the review of empirical studies conducted in different parts of India and
32
studies conducted in abroad respectively. Finally, in section 2.5 we conclude the chapter
of literature review.
2.2. Women’s Empowerment: Issues and the Conceptions
It is widely accepted that empowerment is a quality of human beings and a multi-
dimensional phenomenon. It helps the individual persons or groups to participate and to
get benefit from political or development process in households, communities and
countries. Many researchers and bodies have reported that empowerment is directly
associated with many developmental outcomes. We go to this discussion in details in
section 2.4. Understanding this nexus many developing countries like Nepal, Chile, etc.
have emphasized on empowerment with innovative measures in their Human
Development Reports. However, there is not any universally accepted measure of
empowerment. Nobody would deny that it is difficult to quantify the level of
empowerment. Different studies conducted in different regions have attempted to explain
and quantify empowerment in different ways. As the nature of empowerment is
absolutely influenced by regional culture, different studies have considered different
dimensions and indicators for measuring it. This section has reviewed some important
studies relating to the conceptual framework of women’s empowerment.
The United Nations states that enhancing empowerment means an increase in people’s
ability to bring about change (Human Development Report, 2010). In order to analyse
the components of women’s empowerment UNICEF (1993) has proposed a framework.
These are listed below as cited in Verma (2009).
Welfare: It addresses only the basic needs of women. Women are passive
beneficiaries of the various programmes launched for improvement of their
‘conditions’.
Access: It involves equality of access to resources and opportunities. Women start
recognizing barriers to the accessibility of resources for themselves.
Conscientization: Women are made aware and conscious of gender inequalities as
well as structural and institutional discrimination inherent in the system.
33
Participation: By organizing themselves and working collectively, women get
empowered to gain increased participation and representation in decision making
alongside men equally.
Control: The ultimate level of equality and empowerment, where women are able
to take decision over the various aspects of their lives and play an active role in the
development process. Their contributions are fully recognized and rewarded.
These are actually the stem of the capability approach of development proposed by
Professor Amartya Sen. Thus the expansion of real freedom that the citizens enjoy to
pursue the objectives they have. Obviously, it is a qualitative approach for development.
In different studies women’s empowerment has been defined in different manners.
Keller and Mbwewe (1991, cited in Rowlands, 1995) state that women’s empowerment
is a process which enables the women to organize themselves to increase their self-
reliance, to assert their independent right to make choices and to have control over
resources which will assist in challenging and eliminating their own subordination. The
study of Rowlands (1995) has explored the meaning of empowerment, in the context of
its root-concept: power. According to her empowerment is more than simply opening up
of the access to decision making; it includes the processes that lead people to perceive
themselves as able and entitled to occupy that decision making space, and so overlaps
with the other categories of ‘power to’ and ‘power from within’. She has opined that
empowerment has three dimensions such as Personal, Close relationships and
Collective. The concept of empowerment has been used in different contexts, but it has
been most usefully applied in development context. An empowerment approach centered
on economic activity must pay attention to more than the activity itself. Common view is
that power comes automatically through economic strength. But, she has argued that it
may do, but often it does not, depending on specific relations determined by gender,
culture, class or caste. She has also illustrated that empowerment is a process which
cannot be imposed by outsiders. Although she agrees with the view that appropriate
external support and intervention can expedite and encourage empowerment. In the
context of development Rowlands (1995) has pointed out the importance of individual as
well as collective empowerment. As empowerment is a process where each individual
has to do at her or his own pace, we should take necessary steps for raising level of
34
confidence and self-esteem among poor and marginalized people in such a way that will
enhance their ability to take charge of their own needs. Further, individual empowerment
is one element in achieving empowerment at the collective levels. But concentration on
individuals alone is not enough. She has thought that changes are needed in the
collective abilities of individuals to take charge of their own needs – as households,
communities, organizations, institutions and societies.
Dreze and Sen (1995) have described women empowerment as ability to define self-
interest and choice, and consider woman as not only able but also entitled to make
choices. In order to improve the level of women’s empowerment they have proposed to
reduce gender biasness in mortality rate and natality rates, in access to education and
professional training, in employment, in the ownership of property and in household
work and decision making. Analysing the data from India they have illustrated that
female literacy reduces child mortality rate while both female labour force participation
as well as female literacy reduced female child mortality rate. They have interpreted
these results as evidence of the fact that women’s access to education and employment
had enhanced their ability to exercise agency, i.e., the process of empowerment.
Kishor (1997) has conceptualized empowerment in terms of ‘control’ by which women
would be able to access information, take decision and act in their own interest or for
their dependents. She has considered three categories of composite indicators to measure
women’s empowerment. These are ‘direct evidence of empowerment’, ‘source of
empowerment’ and ‘the settings for empowerment. She has grouped the indicators of
behavioural and attitudinal factors into ten dimensions. We have listed these indicators
including the variables.
Direct Evidence of Empowerment
a) Devaluation of Women: reports of domestic violence; dowry paid at marriage
b) Women’s Emancipation: belief in daughters’ education; freedom of
movement
c) Reported Sharing of Roles and Decision Making: egalitarian gender role;
egalitarian decision-making
35
d) Equality in Marriage: fewer grounds reported for justified divorce by
husband; equality of grounds reported for justified divorce by husband or
wife
e) Financial Autonomy: currently controls her earnings; her earnings as share
household income.
Source of Empowerment
a) Participation in the Modern Sector: index of assets owned; female education
b) Lifetime Exposure to Employment: worked before marriage; controlled
earnings before marriage
Setting Indicators
a) Family Structure Amenable to Empowerment: does not now or previously
live with in-laws
b) Marital Advantage: small age difference between spouses; chose husband
c) Traditional Marriage: large educational difference with husband; did not
choose husband
Mayoux (2000) has clarified basic views regarding the inter-linkage between
microfinance and women’s empowerment. This paper is based basically on secondary
source materials. Following Kabeer (1999) the study has outlined the process of
women’s empowerment considering the analysis of power relation. The power relations
are as follows.
Power within: enabling women to articulate their own aspirations and
strategies for change
Power to: enabling women to develop the necessary skills and access the
necessary resources to achieve their aspirations
Power with: enabling women to examine and articulate their collective
interests, to organize to achieve them and to link with other women’s and
men’s organizations for change
Power over: changing the underlying inequalities in power and resources,
which constrain women’s aspirations and their ability to achieve them
36
These power relations function in socio-economic and political spheres of life at
different levels like individual, household, community, market and institutional.
Reviewing the existing views regarding the impact of microfinance on women’s
empowerment the study has concluded that cost effective ways integrating microfinance
with other empowerment interventions and complementary services are still lacking. In
order to maximize the contribution of microfinance to women’s empowerment Mayoux
(2001) has proposed a strategy, namely, ‘Gender Mainstreaming for Empowerment’ for
donors or Governmental agencies. The core element of this strategy is that gender policy
should integrate productive and reproductive work, welfare concerns and measures to
address power inequalities in strategies for both women and men. Particularly, equality
in access to all microfinance services and an adequate and non-discriminatory regulatory
framework is required for empowering women through microfinance.
According to Kabeer (2001) empowerment refers to the expansion in people’s ability to
make strategic life choices in a context where this ability was previously denied to them.
She has analyzed the concept of women empowerment based on three dimensions
namely, Resources, Agency and Achievements. Resources occupied by the individual
can be materials, social or human which have been treated as conditions of
empowerment. The second dimension of empowerment relates to Agency which acts as
the process of empowerment. According to her agency encompasses a wider range of
purposive actions, including bargaining, negotiation, deception, manipulation, resistance
and cognitive processes of reflection and analysis. Resources and agency together which
Kabeer (2001) refers to as outcome of empowerment, constitutes the potential that
people have for livings the life they want. The ability to choose has been considered as
the central theme of the concept of power. Power may have a negative sense like threat
but empowerment changes the negative senses to a positive one. Empowerment can
reflect change at different levels. At the immediate level empowerment is recognized by
individual resources, agency and achievement. It occurs at the intermediate level, in the
rules and relationship which prevail in the personal and socio-political spheres of life. It
can also occur in the ‘Deeper’ level which changes the distribution of resources and
power in the society and reproduces it over time. She has critically assessed the measures
of women’s empowerment used in different studies. Usually, indicators of the resource
dimension are measured by the access to some specific assets or services. She has rightly
pointed out that many studies fail to consider the differential prior possibility of having
37
access to a particular resource. One should be alert in this line when resource-based
measure of empowerment will come on hand. In order to measure the agency
dimension, she proposes some decision making power of person towards the betterment
of the lives. According to her, in India typical measures of agency dimension include the
decision to purchase of food, purchase of major household goods, purchase of small item
of jewellery, course of action if child falls ill, disciplining the children, children’s
education. To measure the achievement dimension of women’s empowerment, Kabeer
(2001) has agreed with the measures considered in the study of Dreze and Sen (1995),
Kishor (1997) and Becker (1997). But she has pointed out that one needs to segregate
between gender-differentiated achievements which signal differences in values and
preferences and those which draw attention to inequalities in the ability to make choices.
However, she has argued that there is no unique linear model by which one can identify
the causes for women’s disempowerment and alter to create the desired effect. Besides,
she has explained that many of the resources, form of agency and achievements of
women’s empowerment are integral to the broader development goals. With this end in
view, she did not specify causal relationship between resources, agency and
achievements. However, for modelling empowerment we need at least a theoretical
causal consequence among these components, which identifies the directional
relationship. This study has insisted us to study the women’s empowerment in two ways.
On the one way we study the resource and agency dimensions of empowerment by
which we shall construct the empowerment index. On the other hand to realize the
achievement dimension which assesses the impact of empowerment index on some
selected indicators of family welfare.
Viswanathan (2001) has critically assessed the implication of the components of Human
Development Index and Gender development Index in Indian context with special
reference to the state of Karnataka. She has explained that most of the indicators of these
measures underestimated the women’s work and its value in the context of India. This
study has highlighted a set of alternative indicators (asset ownership, access to credit and
environmental degradation) for gender development in India. Application of the Gender
Empowerment Measure (GEM) in India has also been criticized. GEM proposed by
UNDP in Human Development Report 1995 considers the indicators such as earned
income share in professional and managerial jobs and share in parliamentary seats.
Viswanathan (2001) agrees that income confers power, but in many cases in India
38
income earning women enjoy less control over their earnings. Besides, poor women in
India are compelled to earn; it is not a matter of their choice. So a higher workforce
participation rate may imply less instead of more empowerment for women. She has
argued that a large share for women in professional and managerial jobs is hardly reflect
the autonomy and self-reliance of women in India where acute gender biased
differentials in wages and nature of work are the norm of employment. To explain this
argument she has cited the example of Karnataka, where female workforce for the state
has been growing at a faster rate than the male workforce. But disaggregated data
between urban and rural areas indicates that male workforce has shifted from low
income and less skilled jobs in villages to better paying skilled jobs in cities. Greater
participation of women in the workforce seems to be due to vacancies of the shifted male
workforce to the cities. As a result, women are remaining subordinate in terms of the
nature of work and wages. This study also suggests that for assessing the political
participation share of women in local body level is a better norm than that in
parliamentary seats in India. She has pointed out that the pervasive presence of domestic
violence is common in India. Income and apparent status hardly manage to reduce it. So
in order to understand the level of women’s empowerment in developing countries like
India we need to take into account the magnitude of domestic violence against women.
Based on the data of unnatural death of women she has explored that the magnitude of
domestic violence in Karnataka is a serious social problem. She has noticed that despite
the existence of administrative and judicial system, National Human Rights
Commission, the National and State Women’s Commission, a large part of the Indian
women is suffering from the problem of domestic violence. They observed that
‘conspiracy of silence’ that operates in all societies is one of the reasons of domestic
violence and the sufferers treat it as an acceptable adjunct to harmony within the homes.
Another reason is, however, the non availability of the data to disclose the enormity and
prevalence of such violence. Often women in India do not report the problem of violence
to police based on the belief that an outsider should not interfere in marital issue.
However, this study did not explore the nature of relation between empowerment and
violence in India.
Based on the existing theoretical studies regarding the women’s empowerment,
Molhotra, et al. (2002) have come to the conclusion that women’s empowerment is a
complex development concept but not broader than the concept of social inclusion.
39
“Process” and “agency” are two essential elements of women’s empowerment. Women’s
empowerment encompasses a progression, where women act as agent, from one state to
another. For example, it may be an improvement from gender inequality to gender
equality for a specific set of families. The study has reported that major number of
reviewed studies fails to capture the process element of empowerment. According to this
study, empowerment of women may vary from her home to other broad area. So it has
suggested to measure empowerment separately at household level, community level and
at broader arenas. Molhotra, et al. (2002) has proposed six dimensions for each level.
These are Economic Dimensions, Socio-Cultural Dimensions, Familial/ Interpersonal
Dimensions, Political dimensions, Legal Dimensions and Psychological Dimensions.
Different indicators for measuring empowerment have been considered for different
dimensions of each level.
Molhotra, et al. (2002) have reviewed two types empirical studies, namely, studies
considering empowerment as outcome of interest and studies considering impact of
empowerment on other developmental variables. Reviewing a large number of studies
they have concluded that factors such as education, employment, positive marriage
condition and microcredit are influential for women’s empowerment. On the other hand,
a handful of studies show that empowerment has some favourable impact on
contraceptive use, spending on nutrition, child wellbeing and reduced fertility rates. For
measuring women’s empowerment Molhotra, et al. (2002), have some suggestions for
the future researchers. First, we should consider context-specific measures, which reduce
the dependence on proxy measures. Second, in order to capture the process element of
empowerment we should collect data across time dimensions. Third, at a minimum,
quasi-experimental evaluation designs and collection of baseline and end line data must
be considered in implementing programs aimed at empowering women. Fourth, more
interdisciplinary interaction is necessary to develop indicators and approaches that
capture the key elements of women’s empowerment.
The study of Agarwal (2003) has suggested a technological model for empowering rural
women. Women’s employments through technological improvement and participatory
approach are needed to improve their lives. This would ensure a sustainable future for
rural India. Technology model described how scientific and technical interventions could
40
improve the quality of life of women in rural areas. This also shows that the following
factors are crucial for women’s empowerment in rural area.
Proper reorganization of the productive and domestic roles of women
Improvement of women’s empowerment needs facilities like drinking water,
health, sanitation, nutrition, family planning, education and security
Gender integrated participatory technology development is required
Improvement of local women motivator as active “change agent” of technology
through talks and audiovisuals, awareness build up through demonstration and
hands on the job training in relevant field
This study has explained that technological development model along with education,
employment, reduction of socially traditional attitudes i.e. religion, family structures etc.
are responsible factors of women empowerment.
Lillykutty (2003) has explained the relation of education and empowerment of women
with the quality of life. She states that empowerment of women is developing them as
more aware individuals who are politically active, economically productive and
independent and are able to make intelligent decision in matters that affect them and
their nations. A woman is said to be empowered if she has command over wealth,
education, social status, skill etc. and has access to formal financial services. This can be
done by creating an environment in which the distribution of power and resources, the
opportunity to engage in productive work, opportunities to access education, medical
care, and health services can move in favour of women population. If we look into the
definitions of women’s empowerment we find that empowerment is a concept of human
agency- self-efficacy on the one hand and on the other hand it is a process. It indicates
that women’s empowerment has at least three dimensions. She has attempted to establish
a strong positive relation between education and empowerment of women. Education
enhances women’s status in the society and leads to greater input into family and
community decision-making. It provides girls with a basic knowledge of their rights as
individuals and citizens. Knowledge and decision making power can place the women on
a more equal footing with male. Education also provides knowledge and skill especially
in the areas of health, nutrition, sanitation and the environment. Education not only
emancipates women from ignorance, ill treatment and dependence etc, but also
41
empowers them to claim their right to quality of life. Finally, education can serve as a
powerful instrument for individuals to achieve power and status in the society. It is a
source of mobility, equality, and empowerment both at the individual and social level.
Verma (2009) has tried to conceptualise the concepts, needs and context of whole issue
of empowerment and has reviewed the various approaches of empowerment of women.
He has defined empowerment as a process of gaining or accessing control over self and
the resources as well as the ideology which determine power relationships. The process
of empowerment tries to change the existing balance of power in a given context.
Analysing the earlier literature he argued that empowerment is a process of capacity
building and skill development. According to author, empowerment may help the
women to participate more effectively at different socio-political forum. With this end in
view, Verma (2009) has set twin goals of women’s empowerment (a) to challenge
subordination and subjugation; and (b) to transform all the structures, systems and
institutions which in any way, may cause or support gender discrimination and
inequality. In order to achieve these goals she has opined that women need to be
empowered in the contexts of individual, social, economic, physical and psychological
aspects. To suggest the suitable strategy for empowering women in true sense, this study
has presented a critical assessment of the existing alternative approaches and strategies
for women’s development, emancipation and empowerment. It is revealed that most of
the approaches are over-lapping in concept and practice. The discussion of this paper has
come to the conclusion that for empowering women we need the full participation of the
people who are already empowered in the formulation, implementation and evaluation of
the action strategies. Verma (2009) has found that social work approach to women’s
empowerment is suitable in this respect.
2.3. Women’s Empowerment and Welfare: The Impact Studies
The concept of empowerment has been recognized as an instrument of human
development since the inception of the UNDP’s Human Development Report in 1990.
In this section we have planned to review the studies, which have investigated the impact
of women’s empowerment on family as well as social welfare. The studies under this
category have tried to explain the contribution of women’s empowerment in investment
and income generation, in women and child health, in child education and awareness and
social capital, broadly, in family well-being as well as social well-being.
42
Jejeebhoy (1995) has considered five dimensions of empowerment for studying nexus
between reproductive behaviour and women’s empowerment in the developing
countries. These are knowledge autonomy, decision-making autonomy, physical
autonomy, emotional autonomy and economic and social autonomy and self-reliance.
Reviewing many studies conducted across the globe she has established that kinship
structure in the family, education of the women and women’s autonomy, which are the
indicators of empowerment, reduce the fertility rate.
The empirical study of Becker (1997) conducted in Zimbabwe has explored the
implications of women’s empowerment on different set of functioning which Kabeer
(2001) refers to as achievements dimension of empowerment. The functioning that have
been considered in this case are the use of contraception and take-up of pre-natal health
care. In this study women’s empowerment has been measured by an index of women’s
role in decision-making in three important issues. These include the purchase of
household items, the decision to work outside and the number of children to have. In
order to separate the effect of women’s empowerment he has carried out a regression
analysis in two steps based on the primary data. In the first step he has estimated the
effects of some likely determinants of these achievement variables. He has found that
household possessions, the number of survival children, the wife’s employment and
husband’s education have some favourable impact on the use of contraception. Aged
women, rural women and who had polygamous husbands were less likely to use
contraception. The likelihood of receiving pre-natal health care was directly related to
the household possessions index, rural residence, women’s age, education and
employment and husband’s education. In the second step, Becker (1997) incorporated
the women’s empowerment index as an extra explanatory variable to his equations to see
what difference it made. Incorporating the empowerment index he has found that the
goodness fit has improved little for the equation in relation to contraceptive use but
improved significantly for the equation of receiving pre-natal health care. In Zimbabwe
Government has committed to ensure the availability of contraceptives through
community based distribution system for family planning. It makes the ‘contraceptive
use’ a routine behaviour of the women of Zimbabwe. That is why, women’s
empowerment has not significant and separate effect on their contraceptive use. By
contrast, women’s take-up of pre-natal health visit is significantly determined by their
43
role in intra-household decision-making as well as by their education and their
employment status.
In Indonesia Beegle, et al. (1998) have examined the impact of women’s characteristics
on pre-natal care and on the incidence of delivery at hospital. They analysed the data on
about 2000 couples from the 1997-98 Indonesia Family Life Survey. Individual assets,
education and social status of the women have been considered as indicator of her
empowerment. This empirical study has revealed that volume of individual assets, level
of education and social status of the women increases the probability of getting pre-natal
and delivery care of the women in Indonesia.
In an exhaustive study Murthi, et al. (1998) have explored the determinants of three
demographic outcomes such as child mortality rate, the total fertility rate, and the
relative survival chances of male and female in India. For empirical analysis they took
the cross-section data of 296 districts from the census data of 1981. It is found that
female literacy has a negative and statistically significant impact on child mortality and
on male child bias. Male literacy has also a negative effect on child mortality but the
effect of male literacy is smaller than that of female illiteracy and statistically
insignificant. Higher labour force participation exudes the extent of gender bias.
Urbanization and medical facilities have reduces the child mortality and accelerates
gender bias. Poverty in India is positively associated with higher level of child mortality.
However, poverty is not the cause of female disadvantages. A higher proportion of ST
population in district reduces the extent of anti female bias in child survival. This study
has also compared the southern and northern part of India regarding the same three
issues of demography. In southern part child mortality rate is higher and a girl child has a
survival advantages over boys. The empirical analysis reveals that female literacy and
labour force participation are crucial for total fertility. This study has admitted the role of
women agency in mortality, fertility, and gender inequality. The direct proportion of
child health, female literacy, and female labour force participation are likely to be more
conducive to reduce fertility than the indirect intervention based on promising the
general level of economic development.
Pitt, et al. (1998) have estimated the impact of male and female participation in group
based credit programs namely, the Grameen Bank (GB), Bangladesh Rural
44
Advancement Committee (BRAC) and Bangladesh Rural Development Board’s (BRDB)
Rural Development RD-12 microfinance programs in Bangladesh on household
expenditure, on women’s non-land asset, on women’s and men’s labour supply and on
boys’ and girls’ schooling, paying close attention to the issue of endogeneity. They have
used a quasi-experimental survey design and village level fixed-effect method to identify
the effects of credit programs in a limited information likelihood framework. For
estimation they have used a primary data collected from eighty-seven rural Bangladeshi
villages during 1991-92. The participation in group-based credit programs is measured
by the quantity of household’s cumulative borrowing. No doubt this is an indicator of
empowerment in rural Bangladesh. Correcting the problem of endogeneity they have
shown that credit received by women and men have strong positive impact on total
annual household expenditure for each program. The estimated effects of female’s credit
on expenditure, for three programs, are near about two times of the same of male’s
credit. Irrespective of the programs only the effects of female’s credit are statistically
significant. In addition to the expenditure effect, the study has shown that women’s
participation in credit program increases their non-land assets value, whereas men’s
participation does not. Another result indicates that the women’s participation in the
Grameen Bank has a positive significant effect on the women’s labour supply. On the
other hand credit goes to the male and goes to the female significantly reduce the men’s
labour supply. Only the Grameen Bank female’s credit has positive and significant
impact on girls’ school enrolment (see also in Pitt, et al. 1996). Both the male’s and
female’s credit from Grameen Bank and Bangladesh Rural Development Board’s
(BRDB) Rural Development RD-12 have also positive and significant impact on the
boys’ school enrolment. Moreover, these estimated impacts of program participation on
women’s labour supply, on schooling of children and on value of non-land asset holding
are free from endogeneity problem relating to program participation.
Analysing the IFPRI survey data collected from 826 households in Bangladesh, 114
households in Indonesia, 1500 households from the Ethiopia Rural Household survey
(1997) and data on 500 households from the Project for Statistics on Living standards
and Development in South Africa (1998) Quisumbing, et al. (1999) have studied the
impact of women’s empowerment on expenditure shares of food, education, health,
children’s clothing, alcohol/tobacco use and child schooling. They have considered the
value of women’s own assets at the time of marriage as indicator of women’s
45
empowerment. This study has reported that more the resource controlled by women
increases the share of expenditure on child education, but not equally for boys’ and girls’
across the surveyed countries.
Quisumbing, et al. (2000) have also conducted a case study in Bangladesh exclusively
to estimate the impact of wife’s and husband’s empowerment captured by assets at
marriage and current assets on expenditure shares of food, clothing and children’s
education. This study is based on the survey of 826 households residing in 47 villages in
three sites in Bangladesh. The empirical findings show that wife’s assets have some
positive effect on the share of expenditure on children’s clothing and education. Current
assets have a positive effect of on food expenditure share.
We have seen that Kishor (1997) has offered three categories of indicators of women’s
empowerment (refers to section 2.2). Kishor (2000) has investigated the effects of these
indicators on child welfare outcomes, particularly, on infant survival rates and on infant
immunization. This investigation is relied on the data of 3783 women in Egypt who had
a birth in the last five years from the Egypt Demographic Health Survey, 1995-96. The
findings of the multivariate analysis of this study have revealed that the source/setting
indicators of women’s empowerment are more important than the direct evidence of
empowerment in the determination of the outcome variables – infant survival rates and
infant immunization in Egypt. She has found that infant survival rate was lower in the
households where women lived or previously had lived with their in-laws and in the
households where there was a large age and educational difference between spouses.
Women’s employment before her marriage increases the survival rate of her children. It
has been found that mother’s employment and education significantly increase the rate
of child immunization in Egypt. However, children were less likely to have immunized
in the households where mother were under the authority of their parents in-laws.
Besides, among the direct measure of empowerment ‘belief in equality in marriage’ has
a positive and significant effect on children’s survival chance and on the likelihood of
child immunization.
In an empirical study Koenig, et al. (2003) have estimated the impact of women’s
autonomy along with some individual, household and community level variables on the
incidence of domestic violence in Bangladesh. They have illustrated a conceptual
46
framework for the determinants of domestic violence. This study is based on the data set
collected in 1993 under the Family Health Research Project of the ICDDR from two
areas of rural Bangladesh. A total of 9620 sample women aged 15-49 year, has been
considered in this study. Of them 3785 reside at Sirajgonj area and others reside at
Jessore. This study has taken the wife’s report of current physical violence as measure of
domestic violence. It is the dependent and binary variable. Individual and household
socio-economic characteristics includes the wife’s age, religion, landholdings, education
of both the husband and wife, number of living sons and family structure. This study has
included two important individual status namely membership in group lending program
and women’s autonomy. The authors have used the five manifest variables reflecting
three dimensions of women’s autonomy drawing on the work of Jejeebhoy (2000). They
have constructed the women’s autonomy index following the methodology of latent class
analysis. This study also included three community-level variables, constructed through
the aggregation of individual responses at the ‘mouza’ level, a civil administrative unit
corresponding closely to the community level. In order to examine the impact of
community level variables they have considered 179 mouzas through cluster sampling
strategy. These are community-level women's education, community-level savings and
the community-level women's autonomy index. As the domestic violence for each
individual is binary, logistic regression has been used for multivariate analysis. Initially,
they have fitted the logit regression for all observation then they have repeated the model
for each area. This study has revealed that 42% of the sample women have currently
faced domestic violence. A somewhat higher percentage of women in Sirajgonj than in
Jessore reported physical violence. Regression analysis has shown that increased
education, higher socio-economic status, non-Muslim religion, and extended family
residence reduce the incidence of violence. The effects of women’s status on violence
were found to be highly context-specific. In the more culturally conservative area,
Sirajgonj, higher individual-level women’s autonomy and membership in group lending
program increase the risks of violence, and community-level variables were unrelated to
violence. In the less culturally conservative area, Jessore, in contrast, individual-level
women’s status indicators were unrelated to the risk of violence, and community-level
measures of women’s status were associated with significantly lower risks of violence.
However, this study has considered only one indicator of domestic violence. So, it fails
to measure the extent of domestic violence.
47
Thus, we find that the issue of domestic violence has become an issue of concern for all
the Government, policy makers and health workers. There is a perhaps very few studies
which discuss the relation between domestic violence and economic empowerment of
women, particularly, in our study area. In this context, the present study sets three
objectives. First, we study the nature of economic empowerment of the rural women in
the district of Bankura. Second, this study explores the determinants of domestic
violence in the district of Bankura. Third, we would like to examine the impact of
economic empowerment along with other determinants on marginal change in the
incidence of domestic violence. The findings of this study would help the policy makers
to frame appropriate policy for the betterment of women so that they can live safe and
sound lives at home.
Maldonado, et al. (2003), in a study, have explained the role of microfinance program
along with women’s empowerment and other individual and household characteristics on
the child education decision of rural households in Bolivia. They have formulated a
theoretical model of utility and apply the count model to estimate the education gap
(expected education – actual education) of rural child. The women’s empowerment
measures the proportion of the accumulated human capital held by the worker women of
the household. Human capital has been measured by the numbers of years of schooling
accumulated by the workers of the household divided by the number of workers. The
estimation is based on two sets of data one is at municipality level and other is at
national level. This study reveals that microfinance program and women’s empowerment
significantly reduce the education gap of the children in rural Bolivia. However, age, the
position of child compared to siblings, household land holdings and poverty index have
positive and significant impact on education gap.
Ahmed, et al. (2006) have examined the effect of physical violence during pregnancy on
prenatal and early-childhood mortality. For this purpose they have collected data from
2199 women in Uttar Pradesh, India and have used proportions, Logistic Regression
model, Hazards Model to examine the risks for prenatal, neonatal, post-neonatal, and
early-childhood (aged 1–3 years) mortality by mother’s exposure to domestic violence.
They found that 18% of sample women experienced domestic violence during their last
pregnancy. After adjusting other risk factors, they found that mothers who had
experienced domestic violence had higher risks for prenatal and neonatal mortality than
48
the mothers who had not experienced violence. They found no significant associations
between domestic violence and either post-neonatal or early-childhood mortality.
Koenig, et al. (2006) have examined individual and community level influences on
domestic violence in Uttar Pradesh, North India. To conduct this study they have used
the multilevel modelling to explore domestic violence outcomes among a sample of
4520 married men. They found that the individual-level variables like childlessness,
economic pressure, and intergenerational transmission of violence increase the risk of
physical and sexual domestic violence, where as higher socioeconomic position of
households lowers it. They also have shown that a community environment of violent
crime and Community-level norms concerning wife beating were associated with more
risks of both physical and sexual violence and were significantly related only to physical
violence.
Schaedel, et al. (2007) have discussed the role of mother’s empowerment in advancing
the education of their children under the School Family Partnership program (SFP) in
Israel. This study has reported that women’s empowerment indicating the involvement
and familiarity with the SFP program is suitable for student’s achievement. However, the
education of mother is not so important for her ward’s achievement.
The article of Rocca, et al. (2008) have studied the empirical relationships between
women’s empowerment and physical domestic violence among young married women
residing at the slums area of the city of Bangalore. In order to explore the determinants
of domestic violence against women this study has applied the unadjusted and adjusted
multivariable logit regression model. They have found that women in love marriages
contrasted with the women in arranged marriages have more experience of domestic
violence. Women whose families were asked for additional dowry had higher level of
violence. Women who worked before or after marriage were more likely to report
spousal violence. This study has revealed that dowry given at marriage and stable-
occupation of husband reduces the risk of marital violence. It proves that the practice of
dowry is a deep reflection of the many form of gender inequality that women experience.
They have opined that as the practice of dowry is pervasive and routine in many
communities in India, the effectiveness of anti dowry and anti violence laws is limited.
Moreover, this study has shown that participation in SHG increases the probability of
49
having domestic violence. This result indicates that unspoken norm of restricted mobility
of the young women is still pertaining in the Indian society. In addition to the anti dowry
and anti violence laws they have suggested some strategies that mobilize women,
families and communities, to challenge the pervasive acceptance of dowry and to
promote gender equality.
Chowdhury, et. al. (2009) have evaluated pattern of domestic violence pattern in non-
fatal deliberate self-harm (DSH) attempters by analyzing 89 DSH cases admitted at three
Sundarban BPHCs. The authors found that most of DSH attempters were young, female,
low-educated and married. They also found that 69.6% of DSH attempters experienced
more than one form of domestic violence and among female DSH attempters; husband
was responsible for 48.48% cases followed by in-laws for 16.67% cases.
Sarkar (2010) has reported the prevalence, characteristics and reasons of domestic
violence, if any, for adult and adolescent females residing in Singur block of Hooghly
district, West Bengal. The study has found that 23.4% of sample women were exposed to
domestic violence in the past year. Maximum prevalence of domestic violence was
observed among 30-39 years age group, illiterate and unmarried females. This study has
explored that the prevalence of domestic violence was found to be higher among the
Muslims than the Hindus. Prevalence of domestic violence was rampant among the
females who were unmarried or who did not give birth to a child. Majority of the
respondents opined that opportunity of education, being economically productive and
better family income would help them to overcome the problem of domestic violence.
In an empirical study, Janssens (2010) has systematically investigated the quantitative
impact of empowerment program namely, the Mahila Samakhya Program in Bihar, India
on social capital. He has assumed that activities of Mahila Samakhya Program have
increased the nature of agency and empowerment of women. In this study, social capital
refers to the behaviour regarding trust, cooperation and assistance of the households.
Trust has been divided it into two levels, viz. trust in community members and trust in
strangers. The study has constructed the normalized index following factor analysis
based on the arguments against statements relating to trust in community level. Trust in
strangers is quantified putting value one if the respondent disagrees to the statement: “if
you meet a new person from outside the village, you should be very careful to trust this
50
person” and zero otherwise. On the other hand, the issue of cooperation has been
measured by assistance among households and joint action to improve community
infrastructure. The assistance variable is measured as the normalized index constructed
from a factor analysis of the five indicators of assistance. The empirical analysis has
covered three districts in Bihar, namely, Sitamari, Muzaffarpur and Darbhanga. It has
used a set of quasi-experimental data collected from 1991 households. The empirical
findings of this study suggest that Mahila Samakhya increases the level of trust in
community members and in strangers. It substantially increases the level of participation
in collective action, either in school projects or infrastructure for the member
households. The impact of the program is especially large among its target group: the
lower castes and the poorest and least educated households. But it has limited influence
in social assistance among households. This study has also reported that the Mahila
Samakhya program has a bandwagon effects and induce others to join in the activity as
well in the area under study. However, this study did not give any explanation of the
measure of women’s empowerment.
Based on secondary data source Kumar (2011) has investigated the inter-state
disparities in India in the status of women and economic development and its changes
over time. The composite indices for women’s status and for economic development
have been prepared using Principal Component Analysis. This study has reported a high
level of dispersion among the Indian states with respect to the various indicator variables
of women’s status. During the period 1980-90 the disparities have declined for most of
the educational and health status variables. The disparities have declined for most of the
educational and economic status variables over the period 1990-2000. The PCA shows
that educational variables are more important relative to the other variables to influence
the status of women irrespective of the periods under consideration. On the other hand,
urbanization rate, per capita income and the number of factories per lakh of population
have been found to be the dominant factors to explain the nature of economic
development of the states. It has found that Kerala, which occupies tenth/eleventh rank
in accordance to economic development, tops the ranking for women status for each year
under study. This decadal analysis has shown that the states of Uttar Pradesh, Orrisa and
Madhya Pradesh have improved their position among the states over time in the rank of
women’s status. On the other hand, West Bengal, Andhra Pradesh and Gujrat have
shown deterioration in their rankings. Moreover, average ranks of the states for the three
51
years in women’s status and economic development have shown that economically least
developed states are also lower in the ladder of women’s status and vice-versa.
Ray, et al. (2012) have recorded the prevalence of different types of ‘life time’ violence
against the women (VAW) under reproductive age in two urban wards in Siliguri
Municipal Corporation. They have found that more than fifty percent of sample women
had ever faced physical violence. As a consequence of VAW, 54.5% of victims suffers
from mental problem followed by 39.2% were experienced to physical injury. Only 4.9%
has consulted physician and 3.6% have reported to police.
From the literature it is thus clear that women’s empowerment affects a wide range of
household welfare indicators. So, we need to review the literatures that help us
understand the determinants of women’s empowerment. In following section we have
reviewed the studies relating to determinants of women’s empowerment.
2.4. Studies relating to Determinants of Women’s Empowerment
In this section we have discussed the impacts of different socio-economic and
demographic characteristics of women on her empowerment. First, we proceed with the
review of existing empirical literature carried out in different parts in India.
Subsequently, we have presented the studies conducted in abroad relating to the
determinants of women empowerment.
2.4.1. Studies on Women’s Empowerment: The Indian Scenario
Conducting an empirical study in the state of Andhra Pradesh Narasimham (1999) has
assessed the role of awareness generation strategy developed by an NGO, AWARE
(Acronyms for action for Welfare and Awaking in Rural Environment) on women’s
empowerment. The findings generated through participatory observation and interviews,
show that in every area of empowerment such as earnings, education, healthcare status,
access to and ownership of resources including land, decision making power, autonomy
and assertiveness, women, who have been exposed to the awareness generation strategy,
do better than the women of the villages where no awareness had been created. This
study reveals earning of women, degree of consciousness, awareness, education,
confidences, ownership of land and membership of any organization as responsible
factor for women empowerment. However, Narasimham (1999) has described just how
52
to improve empowerment but how much empowerment would be improved is not
answered in her study.
Jejeebhoy (2000) has compared the effect of a range of women’s and household
characteristics on women’s autonomy between the two regionally and culturally different
Indian states, namely, Tamil Nadu and Utter Pradesh. In this study, measures of
women’s autonomy have included four dimensions: (a) role in economic decision-
making, (b) mobility, (c) incidence of domestic violence and (d) access to economic
resources and control over economic resources. For empirical analysis she has conducted
a primary survey and collected data from 1842 women in four districts based on focus
groups analysis. The districts of Coimbatore and Ramnathpuram have been selected
from Tamil Nadu and Meerut and Pratapgarh from Uttar Pradesh. First, this study has
attached a score for each dimension in accordance with the extent of autonomy of the
woman. Finally, she has computed a summary index of women’s autonomy taking mean
of the score indices of each dimension. In order to measure the impact of individual and
household traits on the autonomy index, a linear regression model has been fitted for
each state. This study has revealed that the women in Tamil Nadu are far better in terms
of autonomy than the women in Utter Pradesh. However, determinants of women’s
autonomy varied in the two states. She has found that traditional characteristics – the
number of sons they bore, dowry size and nuclear family type – were directly related
with the autonomy indicators in restrictive culture of Utter Pradesh than they were in the
more egalitarian culture of Tamil Nadu. In Utter Pradesh female employment has also a
positive and significant effect on most of the autonomy indicators but female education
has insignificant effect. However, in Tamil Nadu female employment and education
have strong positive impact on their autonomy.
Jejeebhoy, et al. (2001) have extended the previous study of Jejeebhoy(2000). In order
to examine the influence of religion and region on the women’s autonomy they have
covered ten districts of the state of Punjab in Pakistan in addition to the study area of
Uttar Pradesh and Tamil Nadu in India. In this study initially they present a regression
analysis of the summary index of women’s autonomy for Pakistan, Tamil Nadu and
Uttar Pradesh separately. In this step they have found that the traditional factors such that
co residence with mother-in-law, size of dowry, and age are significant determinants of
women’s autonomy in Uttar Pradesh and Pakistan. In Tamil Nadu, by contrast the only
53
traditional factor, age, plays important role in the determination of women’s autonomy.
Education and wage work status are also significant determinants of autonomy index, but
their impacts are stronger in Tamil Nadu. This study has shown that religion has not any
consistent impact on autonomy index in the two states in India. In the second step
Jejeebhoy, et al. (2001) have pooled the data from the three sites and exclusively
estimated the impact of contextual factors, namely, religion (Hindu/Muslim), country
(Pakistan/India) and region (Northern/southern subcontinent) on autonomy index. This
regression analysis has revealed that only the region i.e. sub continental trait is important
in the determination of women’s autonomy. Religion and nationality are immaterial in
shaping women’s autonomy in the South Asian zone.
Reddy (2002) has examined the process of empowering rural disadvantaged women in
Ranga Reddy district of Andhra Pradesh through self-help efforts, micro-credit, income
generating activities and a range of community infrastructure development activities
undertaken by the NIRD action research project. He has argued that possession of
various power resources, such as, personal assets, wealth, lands, skills, educations,
capabilities, information, knowledge, social status and position held, leadership traits and
maneuverability etc. determine the degree of decision making power; that is called the
empowerment of an individual. Under this research project, SHGs meet the capital
investment need of the women which help them to achieve self-sufficiency. It improves
the decision making power and the leadership power of the members; help in the family
planning attitude; improve the health of the children and mother; protect against the sex
discrimination; prevent atrocities on women etc. Finally, the important component of the
SHG is the micro credit package that is designed by District Rural Development Agency
to benefit the disadvantaged women. The interaction of the women with external
agencies develops the confidence and approachability. It improves the professional and
social skills of the women. Reddy (2002) suggests that participation in SHG, which
enhances their institutional and managerial capacity, is an imperative element in the
empowerment process of the rural women. This research study has also shown that the
women, who have significant control over credit provided to them, influence the degree
of empowerment.
According to Varadarajan, et al. (2002) women empowerment is not a question of
giving or providing some provisions, but it is a question of all women being able to use
54
those provisions without hesitation. Women’s empowerment includes two components,
namely, attitudinal empowerment and material advancement. The key area in the
empowerment of women is the economic area. They suggested that the development of
micro enterprises in general and particularly by the women would be appropriate
approach to fight against poverty at the grassroots level and generate income at the
household level. They have explained that self-employment and entrepreneurship create
economic independence and better social status and sometimes it is necessary for
improving their position not only in their family but also in the society where people
treat them with respect. In fact the best non-controversial way to empower women is the
spirit of entrepreneurship.
Evaluating the first phase impact of Maharashtra Rural Credit Program (MRCP), Bhide
(2003) observes that the SHGs have been able to bring a large section of village
populations, who were deprived of the banking service, under the shelter of banking
service. This study has looked at the various financial and social issues related to SHGs.
In order to evaluate the impact of SHGs, Bhide (2003) has considered a sample of 358
SHG members drawn across 147 SHGs in four districts of Maharashtra. This empirical
study has found that SHGs catalyze the savings rate and increase the ability to absorb
credit to create new assets by rural poor and open opportunities for rural poor. The
investigation reveals that SHG as a social organization conducts a meeting per month. In
these meetings the members of the group discuss several social issues like importance of
education, decision-making power within family and in society, family planning,
involvement of the group in village affairs and gender awareness etc. The meetings and
activities of the members of SHGs help build new leaders and new way of collective
functioning in rural area. The participants of MRCP have reported that the formation of
SHGs through MRCP increases the number of credit resource, slightly increases the
income of the members as well as expenditure on education, health etc. and help poor
people to shift towards self-employment. They have also reported that SHG programme
could not change the power of mobility or household decision-making power of women.
It helps the rural poor women to enhance the confidence level.
Jain, et al. (2003) have statistically analyzed the impact of membership of SHG on
various dimension of women empowerment. To estimate the impact they have used a
primary data collected from Kunpur Dehat district of Uttar Pradesh. They have found
55
that the membership of SHG enhance the quality of status of women as participants,
decision makers and beneficiaries in the democratic, economic, social and cultural
spheres of life. Their statistical result shows that the number of total SHGs members had
increased their level of education than those of non-member of SHGs. The SHGs
members had higher housing facilities than non-members. The number of total SHGs
members had increased their exposure of mass media, extension orientation,
occupational level, size of holding material possession, annual income, membership of
the organization are significantly higher than those of non-members of SHG. This study
confirms the effectiveness of membership of SHG to improve women’s empowerment.
In a micro study conducted in the foothills of Himalayas, Handy, et al. (2004) have
explained the role of NGOs in empowering woman in rural India. They have constructed
an empowerment index including four dimensions of empowerment, namely, personal
autonomy power, family decision making power, economic and domestic consultation
power and political autonomy power. To do this they first measured an index for each
dimension; then aggregating these indices they constructed the empowerment index
which they have called E-index (empowerment index). To document the level of
empowerment among women who participated in NGOs, they have collected data from
the different employee groups. At the leadership group they have interviewed
supervisors of various programmes. At the follower level they have interviewed
fieldworkers who went into the village and worked directly with village women. They
also choose to interview some local woman living in the area that the NGO served.
These women are called recipients. Measuring the E-index for each group they have
found that empowerment levels of different groups are significantly different.
Particularly the E-index of supervisors is higher compared to that of field workers and
recipients. To account for this difference they search the impact of various socio-
economic-demographic factors on the E-index. Because they observed that socio-
economic-demographic status of different group are different. From the literature survey
and their personal observation they have expected that the E-index is a function of age,
family structure, income class, education and tenures at NGO. In order to understand the
combine effects of all conceptualized variables they did not include the recipient group,
as they all had zero years at the NGO. They have found that the years of participation in
the NGO and education level of woman are significant explanatory variables of women’s
empowerment but income class and family structure are not.
56
Sridevi (2005) has studied the empowerment of lady teacher of the schools in the city
Chennai, India. She has analyzed the empowerment in two ways, one is qualitative
analysis and other is quantitative analysis. For two parts she has considered ten
determinant of women empowerment. These are family type, age, educational
qualification, household time spend, freedom of mobility, spouse age differences,
educational differences with spouse, number of children, control over personal salary
and supporting the natal home. In both types of analysis these factors are taken as
independent variables. She has derived the empowerment of women in the form of a
qualitative variable. This is obtained from the respondents’ perception about whether
they feel empowered or not. If a respondent feels empowered then the value is 1 and 0
otherwise by this manner the dependent variable was derived in the form of binary
variable. Then applying logit model she has analyzed the significant impact of the above
independent variable on the log odds ratio of women’s empowerment.
In the next part for quantitative analysis, she has constructed empowerment index of
women using a weighted average. To calculate the empowerment index she has used
various proxy variables. These are (a) control over personal salary, (b) maintenance of
family income, (c) supporting the natal home (d) financial decision on own health care
and (e) expenditure on the education of children. In this step she has taken empowerment
index as the dependent variable. Then, by using a multiple regression analysis, she has
found the direction and the magnitude of each factor’s influence on the empowerment of
women. For both type of analysis she has used a personal stratified random sample of
eighty women, postgraduate school teacher, collected from the city of Chennai in India.
From the both types of analysis she has found more or less same result. The study
reveals that age, household time spend on household work, freedom of mobility, spousal
age differences, educational differences with spouse, control over personal salary and
supporting the natal home are significant determinants of women’s empowerment in
both type of analysis. However, educational qualification is significant only in
quantitative analysis.
The study of Vasimalai, et al. (2007) have attempted to focus on the principles and
socio-economic impact of the Kalanjiam model of group based microfinance. This
program is based on the principles of self-help, mutuality and ownership by poor
women. On the basis of a random sample of 300 respondents, of which 240 belong to
57
Kalanjiam model, Vasimalai, et al. (2007) has assessed the impact of model on economic
development of the family and on empowering the marginalized group of women. The
study has shown that the family income of the members has increased at an increasing
rate with the age of groups. Comparing the empowerment of women member of
Kalanjiam model and control grouped women the study has established that women’s
empowerment of the Kalanjiam member is greater than that of the non-members. To
make this comparison authors have considered domestic violence, sharing the household
works by the husband, assets owned by the women, decision making ability, personal
skills, women’s share in household income, exposure to outside of the family, interaction
with the community, leadership ability and women’s space in the public sector as
criteria.
Meher (2007) has assessed the impact of SHG based microfinance on income poverty
and empowerment in KBK region of Orissa collecting information from 77 members of
selected five women SHGs. The study reveals that SHG based microfinance is successful
to reduce income poverty. He has considered the variables like importance in family,
awareness regarding education and health care as indicators of the social domain of
women’s empowerment. Economic domain is explained through the indicators like
economic self-sufficiency, consumption of nutritious food, purchase of consumer
durables and awareness regarding the use of utensils. Participations in gram sabha and
political awareness have been taken as indicators of political domain of women’s
empowerment. The positive change of a particular indicator indicates the incremental
improvement of the corresponding domain and women’s empowerment as a whole.
According to Meher (2007) positive change of an indicator up to 40% due to
participation in SHG indicates low impact on empowerment. Above 60% positive
change of the indicator implies high impact and 40-60% positive change implies
moderate impact of the participation in SHG on empowerment. Based on this research
design the study has informed that members of four SHGs have scored moderate impact
and members of remaining one group have scored low impact with respect to social
empowerment. In terms of economic domain of women’s empowerment the members of
all the groups under consideration fall in the low impact category. Sample SHGs have
moderate impact on the political empowerment level of the members. The study has
reported that all the SHGs, except one, have low impact on women’s empowerment
considering the average impact of the three domains. The exception one has shown a
58
moderate impact of empowerment. Indeed, the study has shown a gloomy picture of the
performances of SHGs in the process of empowerment generation among the group
members.
Chattarjee (2008) has examined the impacts of SHGs on income, employment and
empowerment status of women in Khejuri block of Purba Midnapur district, West
Bengal. This study has shown that SHGs generate income and employment of women
members at an admirable level. In order to assess the role of SHGs in empowering
women he has considered six elements of empowerment, namely, importance in family,
role in deciding the number of children, decision making power in family matters,
increase in self dependence, securing the respect of husband and in-laws and decrease in
domestic violence. The findings of the study are as follows. As the women earn more
through SHG, the importance of them in family increases compared to unemployed.
Self-employed women have more freedom in deciding the number of children. SHGs
members play a dominant role in taking decision on different matters such as saving and
expenditure in family; education of children; pattern of consumption. Members having
higher income enjoy higher level of self-dependence in family. Economic emancipation
and engagement in broader social system have reduced domestic violence against
women. SHGs have inculcated conscious in women regarding health and education for
their children. Almost all the respondents have completed the immunization package or
been continuing this in due course.
Nayak, et al. (2009) have analyzed the status of women empowerment in India using the
data of NFHS-3 (Government of India, 2005-06). In this study decision making power of
woman in household, freedom of movement, acceptance of unequal gender role, access
to education, access to employment, exposure to media, experience of domestic violence
and political participation have been considered as indicators of women’s empowerment.
This study reveals that the decision-making power of woman in household varies
directly with their age, education, and husband’s education level. Employed women are
likely to have more decision-making power in household than the unemployed women.
In urban area and in nuclear type of family women enjoy more autonomy in household
decision-making. It is found that mobility of widow or divorcee is more than ever
married women. More than half of the sample women believe and accept intra household
unequal gender norm against women. This attitude does not vary significantly with age
59
or structure of family but declines sharply with education and for urban region of
residence. The study has shown that participation of girls at all stages of education has
been increasing overtime. Not only women lag behind men in terms of access to
employment, but also a major portion of employed women does not have full control
over their cash earnings. However, control over their cash earnings is positively related
with age and place of residence and education, but not varies significantly with
household structure. In terms of exposed to media women are disempowered relative to
men in India. Extent of domestic violence is lower in urban areas as compared to rural
areas. Experience of domestic violence of ever-married women is higher than that for
never married women. Aged women are more victimized of domestic violence compared
to younger. Women are also less empowered in terms of casting votes and representation
in general elections overtime. Nayak et al. (2009) has also identified several constraints
in achieving desired level of women’s empowerment. These are poverty, social norms
and family structure, lack of awareness about legal and constitutional provision etc.
Samanta (2009) has examined the effectiveness of SGSY-centric microfinance
programme on women’s empowerment in Burdwan district, West Bengal. She has
considered the six indicators of empowerment, namely, mobility, confidence and
capacity building, entitlement, perception of empowerment, decision making, and
autonomy and authority. The study has revealed that the women have improved their
capacity of financial management and can participate in financial decision of the family
after joining SHGs. Although in most cases income of the women are spend for
maintenances of family, husband or other family members, hardly allow them to keep
the money with themselves. In this study, 72% of the women have reported that they
have been empowered after joining SHGs, but almost half of them could not clarify why
they feel empowered. A section of the sample women have felt that they have achieved
the power of bargaining for their own well being. Samanta (2009) has reported that
sample SHG member women enjoy significant authority regarding voting in election and
children’s education and very little independent authority in family planning, family
expenditure, going outside and medical treatment.
In an empirical study conducted in Burdwan district in West Bengal Adhikary, et al.
(2011) have computed the degree of women’s empowerment considering five
dimensions, namely, economic, socio-cultural, familial, political and legal dimensions.
60
This study has revealed participation in Self-Help group, education level of the women,
time spent on unpaid household job, type of family, nature of residential region and caste
as important factors affecting the degree of women’s empowerment.
A study of Adhikary, et al. (2011) based on 964 households’ data has shown that
compared to men, women are likely to have less access to formal credit. However, the
participation in SHG ensures the accessibility to formal credit for the rural women and
increases the household expenditure on food and nutrition, on fuel and energy, on health
care and on education for the rural poor households in the district of Bankura, West
Bengal.
Conducting a field survey in Hosakote in Bangalore (Rural) District, Karnataka
Anuradha (2012) has analysed the changes of women’s empowerment after joining self-
help groups. This study has applied PCA to construct empowerment index and observed
socially viable, personality outlook, economically strong, living standard and
accessibility as important component of women’s empowerment. Reorganizations in
community, literacy level, access to credit and health, voicing concern have been taken
under socially viable variable. It is reported that SHG membership facilitate the women
various opportunities to involve in various activities which empower them by enhancing
their role in the society. This study has grouped nutrition awareness, decision making
related to child centred and money centered, participation in development programme
and increase in confidence level under personality outlook. It has found that there has
been a positive transformation in personality outlook after joining SHGs. The author has
named ownership of house and land and improved relation with husband as
economically strong. This study shows that economic position of women has improved
significantly after joining SHGs. The variables, change in personal financial position,
change in share in family income and reorganization in family have been grouped under
the living standard. This study shows that SHGs have improved the living standard of
women. Better interactions with outsider, access to credit and asset building have been
levelled as accessibility. The association among the SHG members and officials of
different offices help the SHG members have connection with outsiders. Participation in
SHG has also ensured access to credit and other financial products.
61
2.4.2. Studies on Women’s Empowerment: The Global Scenario
In the new millennium enhancement of women’s empowerment has got importance
across the globe as process of economic development. Many theoretical and empirical
studies have been conducted in different parts of world outside India under the discipline
of Development Economics. In this section we are going to review some selected
studies, conducted outside India.
Malhotra, et al. (1997) have empirically analyzed how education and employment
status of women, other life course and household factors affect women’s empowerment
in Kalutara district of Sri Lanka. They have considered whether the woman has most
control over the money matters and other social and institutional matters in the family as
measure of empowerment. In order to measure the impact of education, employment and
other factors on empowerment the logistic regression analysis has been used. The study
has revealed that years of schooling and current employment status of women have
positive and significant impact on the financial decision making power of women in the
family. But, after a certain level, additional schooling does not contribute to increased
control of financial matters. However, education and employment are immaterial to have
power on social and organizational issues. Past work experience is an important
determinant of the financial decision and of the social/organizational decision-making
power of the currently married women. The study has shown that poor women lay
behind the women from middle-income group families in terms of enjoying the power of
financial and social/organizational control in the family. The empirical study has
reported that the women from Moor families are less likely to have control over financial
and social/organizational decisions than the women from Sinhalese families. It is also
found that the motherhood status of women increases her social and organizational
power but not her financial power. Women of large families and those residing with their
parents are better in position to have input in social and organizational matters. However,
husband’s characteristics don’t have significant effect on woman’s decision-making
power in the family in Sri Lanka.
Amin, et al. (1998) have explored the relationship between poor women’s participation
in NGO based microcredit programme and their empowerment using empirical data from
rural Bangladesh. In order to quantify women’s empowerment this study has constructed
three indices, namely, inter-spouse consultation index, individual autonomy index and
62
authority index. It has been found that NGO credit member enjoys higher level of
empowerment than the non-member, irrespective of non-member’s residence in
programme area or non-programme area. The non-member in programme area shows a
higher level of empowerment on the autonomy and authority index than do the non-
member of non-programme area. Education, house type, yearly income, duration of
NGO credit membership and non-agricultural occupation are positively related with
autonomy and authority index. Both the indices vary with age and region. The level of
empowerment for theses indices vary directly with age. The regions having traditional
restrictions on women and less coverage of the NGOs have lower empowerment level of
women. Membership of NGO and residence in an NGO programme area are
significantly and positively related to autonomy and authority index. Income has positive
relation with consultation index. Concrete or corrugated buildings, areas of residence
outside the southern and eastern region, non-agricultural occupation, respondent’s
education and respondent’s age have significant and positive relation with either
autonomy index or authority index. NGO credit membership has strongest effect in
explaining the variation in women’s empowerment. Focus group discussion reflects that
the relation between NGO credit membership and indices of women’s empowerment is
due to increased contribution to the economic survival of their families, due to their
credit related movement outside their home and due to their participation of NGO
consciousness raising activities.
Arends-Kuenning, et al. (2001) have examined the rural Bangladeshi people’s view
about the benefits of education of women. For this purpose they have used data from in-
depth interview conducted in 1996 and 2000 in two villages located in the Rajshahi
district of Bangladesh. From the in-depth interview, the authors have realized that
parents think that daughter’s well-being is best secured through marriage and education
is very much valuable in marriage market because it is an input in children’s education.
From the in-depth interview the authors have observed that education helps women to
produce human capital, to enhance women’s income earning power and to increase
women’s bargaining power and to catch respect within the family. With the spread of
micro-credit, targeting the poor women, people have recognized the importance of
education for managing micro enterprises. All these are different dimensions of women’s
empowerment. Hence from this study it is obvious that education accelerates the
empowerment level of women. Thus, we can conclude that education is a determining
63
factor of women’s empowerment. This study does not provide any statistical analysis
and does not directly say anything about the impact of education on empowerment.
Mason, et al. (2003) have viewed women’s empowerment as a part of sociological and
anthropological conception. They have defined domestic empowerment of women as
their freedom from being controlled by other family member and their ability to affect
desired outcome within the household. They have considered four aspects of domestic
empowerment namely economic decision-making power, family size decision-making
power, physical freedom of movement, husband control on them via intimidation and
force. Using a household level data collected from 56 communities in Pakistan, India,
Malaysia, Thailand, Philippines in 1993 and 1994 they have shown that community
differences are significant in measuring women’s empowerment. This study has
established the fact that community and country as opposed to individual and household
characteristics are able to explain more variation in empowerment i.e., a women’s
community can better explain her score in particular empowerment measure than can her
own age, education, age at first union or economic experience.
Ghuman, et al. (2004) have illustrated the nature of gender relations and difficulties in
its measurement using data collected from 23 communities in five Asian countries,
namely, India, Pakistan, Malaysia, Philippines and Thailand. They have asked the same
questions to the married women ages 15-39 years and to their husbands at different times
regarding the women’s autonomy. For measuring women’s autonomy they have focused
on the Freedom of Movement, Decision-Making regarding Children, Household Tasks
and Decisions as domains of women’s autonomy. Applying an item response model the
study reveals that wives and husbands have significantly different assessment on
women’s autonomy in various domains. They are also different in terms of cognitive
understanding of the responses regarding the focused domains. The nature of these
differences also varies across the contexts and across the communities. The study has
found that husbands in South Asian communities ascribe higher autonomy of their wives
than wives do for themselves. Finally, Ghuman, et al. (2004) have estimated a logit
model in order to measure the impact of differing perspectives of wives and husbands
regarding the women’s autonomy on the experience of infant mortality. It has been
found that women’s autonomy has negative impact on the child mortality if the woman
reports on her own decision making power on what to do when the child falls sick. On
64
the other hand, if the husband reports that his wife takes decision regarding the matter; it
has positive impact on child mortality in the sample countries except Pakistan. However,
Ghuman, et al. (2004) have opined that these survey questions are of limited utility for
understanding differences in gender stratification across different contexts.
Conducting an empirical study in rural Vietnam Santilan, et al. (2004) have assessed
nature of women’s empowerment in the socio-economic sphere as well as in
reproductive health. For this study they have interviewed 57 married women between the
age of 22 and 52 years and 13 of the women’s husband. They have conducted a
qualitative analysis of the case study material assembled from the in-depth interviews.
This study has developed two sets of domains along with their indicators for women’s
empowerment. One set is related to socio-economic sphere and other is related to
reproductive health. The domain of socio-economic empowerment includes production,
housework, family expenditure, relation with natal relatives, community participation
and right of husband and wife in the family. The reproductive health domain includes
decision making about childbearing, contraception, sexual communication and
negotiation, pregnancy, appraisal of health services reproduction tract infection,
reproductive health role and right. Each indicator has been attached a score ‘one’ for the
responses that has little or no evidence for empowerment, ‘two’ for moderate evidence
and ‘three’ for considerably empowered respectively. Each woman receives a score for
each indicator and an overall score for each domain. They have pointed out the
methodological challenges that they faced in analyzing women’s empowerment. They
have faced challenges in conceptualization of empowerment. They have had conscious
regarding the overlapping concepts. To avoid the ‘politically correct’ response they have
made a cross-checking of the response from community leaders. In order to collect the
responses in a better way and for collecting the appropriate answer of the sensitive issues
they have deployed systematic analytical framework to open ended data and competent
surveyors. Based on the field survey, they have reported that the women exert a
significant degree of control in daily decision making about productive activities.
Regarding the family expenditures usually women take decision jointly with their
husbands, but in case of disagreement husband enjoys more power. A major portion of
the sample women participates in community occasions. Many respondents agree with
the one or two child policy. However this study is too small to draw definite conclusion
65
regarding the nature of women’s empowerment in Vietnam. But it provides an
invaluable application in further study of women’s empowerment.
Parveen, et al. (2004) have measured and estimated the empowerment of the women
resided at three villages of Mymensingh district of Bangladesh. They have constructed
cumulative empowerment index taking six empowerment indicators, namely,
contribution to household income, ownership of assets, access to resources, participation
in household decision-making, perception on gender awareness, coping capacity to
household shocks. They have considered seven influential factors - two at individual
level, two at household level and three at social level as determinants of women’s
empowerment. These are formal education, non-formal education, sexes of children,
spousal relationship, media exposure, spatial mobility and traditional socio-cultural
norms. The regression analysis has revealed that formal and non-formal education, sexes
of children, spousal relationship, exposure to media and spatial mobility have positive
and highly significant effects on cumulative empowerment index. On the other hand,
traditional socio-cultural norms have a significant and negative impact on cumulative
empowerment index.
Parvin, et al. (2004) have examined the performance of income generating activities
supported by Rural Women Employment Creation Project (RWECP) on empowerment
of poor women in Dumuria Thana of Khulna district in Bangladesh. The study is based
on a set of primary data collected from a random sample of eighty women who are
members of RWECP. In the study they have considered three indicators of women’s
empowerment, namely, participation in household decision-making, control over income
and access to assets. In order to analyze the impact of income generating activities under
RWECP on women’s empowerment the study has used descriptive statistics and
Weighted Mean Index (WMI) method. The study reveals that engagement of women in
income generating activities under RWECP has enhanced their capability to express their
opinion and make decision to meet personal needs, to contribute to buy households
assets, availing treatment and recreational facilities independently. However, in most of
the cases income of women are controlled by their husbands. Findings show that widows
and abandoned women are in better position to control their earnings than the women
living with husbands. Low level of control over income does not allow them to have
significant access to assets of their own. The constructed WMI of empowerment has
66
shown that the level of empowerment is noticeably high for the women engaging in cattle
rearing and low for the women engaging leather goods making. The study has concluded
that marital status and religion are influential factors of women empowerment in rural
Bangladesh.
Williams (2005) has defined women’s empowerment as relative level of adherence to
current and context specific gendered norms. He has identified six dimensions of
women’s empowerment, namely, sense of self and vision of the future, mobility and
visibility, economic security, decision-making power in the households, participation in
non-family group, interact effectively in public sphere. This study has formulated a
conceptual model of women’s empowerment. In this model he has shown that economic
and gender components of empowerment influence each other and both have interactive
effect on exercise of power that effect health, fertility, mortality etc. He has further
shown that economic resources available to women have independent influences on
these demographic achievements. He has used confirmatory factor analysis to test the
hypothesis, women’s empowerment is multi-dimensional and whether the indicators
chosen conform to theoretical model. Calculating the goodness of fit index, Tucker-
Liews index, Comparative fit index he has found that six dimensions model fits
significantly better than one dimension model. He has estimated correlation and variance
between two components of latent dimensions. Some of these correlations are
statistically significant. Thus he has shown that empowerment is multidimensional in
nature and theses dimensions are interrelated.
Moser, et al. (2005) have explored the success and limitations of gender-mainstreaming
policies of international development institution. The common policies include six
components- a dual strategy of mainstreaming gender combined with targeted actions
gender equality, gender analysis, a combined access to responsibility, gender training,
support to women’s decision-making and empowerment, monitoring and evaluation.
Besides, work with other organization, budget, and knowledge resources have been
considered by some of the organizations. However, they have observed that the
implementations of the gender mainstreaming policies are inconsistent and involving in
few activities, rather than coherent and integrated process. The implementation of
policies consists of institutional and operational inputs, which are closely interrelated.
The most of gender mainstreaming evaluations focus on institutional inputs rather than
67
the process of operational and programming implementation. They have identified a
range of constraints regarding implementation. First, the lack of responsibility,
commitments and skill of gender specialist are primary constraint regarding program
implementation. Second one is the male leadership which discriminates female staffing
in term of attitudes, recruitment, working conditions, structure and procedures.
Sometimes it excludes women by scarcity of high level of job shares, extensive travel
requirements and long work hours that are difficult to do for women with dependent
children. Third constraint is the voluntary accountability. Gender training is another
constraint in implementing gender policies. Constraints may also come from operational
aspects of gender mainstreaming. The common operational constraints are the lack of
effective, consistent and systematic monitoring and mainstreaming outcomes and
impacts. Terms of women’s participation in economic activity have been identified as
another operational constraint.
Conducting an empirical study in Sylhet district of Bangladesh Hossain, et al. (2006)
have argued that change approach is most suitable for empowerment of women in
Bangladesh. The ‘change approach’ includes integrated development, economic
empowerment and consciousness raising approach and the changing of the attitude of
male towards female. In this study empowerment means women’s authority to make
choices and decision that facilitates the development of knowledge and control over
resources to exercise the right. This study explores that women in Bangladesh have to
rely on their male guardian’s opinion in taking decision. In this study 79% of women
reports that they are not usually congratulate any income generating activities in case of
husband disagreement. Though nearly 62% women of this study have self-income but
more than 65% says that their participation in decision-making remains unchanged in
spite of increasing their income. To increase the consciousness of the women,
development workers have taken several awareness campaigning. However, 52% can’t
go outside their home conniving at their husband opinion. Their participation would be
ensured at the meeting if their husband allows them for it. Most of the sample women
have identified the male dominance attitude as a major hindrance to empower women.
They have expressed their opinion in support of changing the attitude of male. About
two-third of the sample women think that ‘change approach’ which covers all the
changes in the society related to women’s empowerment, i.e., changes in values and
attitudes, income, employment, education, access to property and resources, participation
68
in taking decision in the family, is most important and most effective for women’s
empowerment.
Bhuiyan, et al. (2007) have defined empowerment as a mechanism of awareness and
capacity building leading to greater participation in decision-making and greater access
to and control over physical resources as well as power structure. This study says that
empowerment includes women’s participation in work place. The authors narrate that the
entrepreneur development focuses on women’s empowerment in general and their
participation in income generating activities in particular. Women’s entrepreneur help
women increase the courage to talk with people associated with business and trade. They
say that entrepreneur increases women’s participation in decision making in family and
social matter; removes social seculation and religious sanction against women. It
decreases social discrimination against women and increases the income of the women.
The authors have argued that women’s entrepreneur changes the social-psychophysical
quality of individual women, which is most effective for empowerment. Working as an
entrepreneur improves the woman’s social and economic status. Trade and generation of
income increase the self-confidence of women. Finally we can conclude from the study
that working as entrepreneur increases the empowerment of women. But, working as
entrepreneur is not the only determinants of women’s empowerment, there are many
other determinants of women’s empowerment. This study remains silent about the other
determinants of women’s empowerment.
Mostafa, et al. (2008) have tried to measure the women’s empowerment index (WEI)
score for Bangladeshi women using Bangladesh demographic and health survey
(BDHS), 2004, data. They measured WEI in domestic sphere using three dimensions-
women’s economic decision-making power, household decision-making power, and
physical freedom of movement. Each dimension has some relevant indicators and on the
basis of this indictors score was given to each dimension. They have concluded that
though the level of women’s empowerment is not satisfactory for any age group, older
women have more independence and empowerment than younger.
Chowdhury, et al. (2009) have explored the key determinants of women’s
empowerment in a remote area of Pakistan. In order to carry out this study the authors
collected the primary data from southern Punjab from 200 respondents using stratified
69
random sampling techniques. All the respondents are female of age group 14-65 years.
For measuring women’s empowerment they constructed cumulative women’s
empowerment index (CWEI). CWEI is a composite index of four separate indices,
namely, personal autonomy index, family decision-making index, economic domestic
consultation index and political autonomy index. They have considered independent
variables like education of women, years of women’s schooling, doing any paid work,
having bank account, access to media, opportunities for outing, per capita income of the
households, participation rate dependency ratio, education index of households, age of
women, marital status of women, asset ownership by respondent, knowledge of Islamic
women’s empowerment, joint family structure, living in community, not believing on
typical out-dated socio-cultural norms, number of household member, fear of violence
from father/husband and distance of health unit from respondent’s home. For analyzing
data, they used descriptive analysis and ordinary least square. They have considered four
regressions analysis- one for total sample, one for urban area, one for rural area and one
for tribal area. The regression analysis based on total sample shows that woman doing
paid work, having bank account, access to media, opportunities for outing, age of
women, married women and knowledge of Islamic view point of women’s
empowerment have significant and positive impact on women’s empowerment. Women
doing paid work, participation rates and age of the woman have positive impact on
women empowerment in rural area. Married women and doing paid work have
significant positive impact on empowerment of the women belonging in tribal
community.
Ashraf, et al. (2010) have explored the impact of a commitment savings product on
female empowerment in Philippines. They have considered the special savings product,
called SEED (Save, Earn and Enjoy Deposits) account with the Green bank of Caraga, a
small rural bank in Mindanaano, Philippines. It is a term deposit and individual savings
account. Female empowerment has been quantified in this study by constructing two
decision making indices from nine decision making situations. The nine situations refer
to decisions on what to buy at the market, expensive purchases, giving assistance to
family members, family purchases, recreational use of the money, personal use of the
money, number of children, schooling of children and use of family planning. For each
decision it assigned value two if the respondent exclusively takes the decision, zero if the
spouse takes the decision and one if both take it. The first empowerment index is
70
constructed taking equally weighted mean of each response of the nine decisions. The
second index is the linear combination, determine through a factor analysis, of the
individual responses to each questions. For empirical analysis they have randomly
chosen 3125 adult clients of the Green Bank. They have also made a randomization to
divide the total sample in to three groups. The first one is the commitment treatment
group, who were counselled regarding the importance of savings and offered SEED
product, second one is the marketing treatment group who were counselled but not
offered SEED product and the third is control group who were neither counselled nor
offered SEED product. This study reveals that assignment to the treatment group
strongly increases the both decision making indices for married women but not for
married men. Not only that, the average effect is largely driven by increases in decision
making ability for women who were below the baseline median. The marketing has a
smaller effect on changes in decision making indices. This study has also examined the
effect of the SEED product on household expenditure pattern on durables and savings
attitudes. It is found that SEED product increases the purchase of consumer’s durables
associated with female use. Ashraf et al. (2009) has argued that it is happened due to
empowerment effect of the SEED product. Indeed, the opening of SEED account has
improved the savings practices of the treatment assigned bank clients.
Varghese (2011) has conducted an empirical study on women’s empowerment in Sohar
region, Sultanate of Oman. This empirical study is based on a set of primary data
collected from 150 women. This study has measured the women’s empowerment by
identifying the household decision making ability, assessing economic decision making
capability and evaluating the freedom of mobility of the women. According to the author
there are the three dimensions of women’s empowerment, namely, economic,
households and social. For each dimension, an index of women’s empowerment has
been computed following the methodology of UNDP used in the calculation Human
Development Index. Finally by taking the simple average of these dimensional indices
the author computed the Women Empowerment Index (WEI). The computed
dimensional index has shown that women of Oman are forward in terms of household
and economic empowerment but they are backward in position in terms of social
empowerment. In order to quest the responsible factors affecting women’s empowerment
this study has considered five socio-economic statuses of women. These are income,
education, employment, acquisition of assets and media exposure. For statistical analysis
71
she has applied ANOVA and regression technique. This study has found that income,
education, employment, acquisition of assets have positive and significant impact on
women’s empowerment whereas the media exposure is insignificant in the determination
of women’s empowerment. Finally, this study has taken women’s perception on their
empowerment. For this purpose, this study has considered three issues (the right to
protection against harms, the right to set up association and the right to get own land)
relating to legal rights and one issue (the right to involve in country’s politics) relating to
political right. The primary survey of this study reveals that majority of the sample
women are aware about the right to protection against harms. Only 29 percent of the
sample women have strongly reported that they have the right to set up association.
Nearly 20 per cent are disagree or neutral regarding their right to get own land. However,
almost half of the sample women have opined that they do not have right to involve in
country’s politics. So far, this study has considered only the women’s perception
regarding their empowerment. For perception analysis one should consider the
perception of the person concerned as well as other persons within the family.
Zaman, et al. (2012) find out the level of empowerment of women in household
decision making process in some purposively selected rural and urban areas of
Bangladesh. It has considered 18 variables in household domain, namely, freedom of
purchase saris, freedom of purchasing cosmetics, opinion for children’s admission in
school, opinion seeks for child birth, opinion for family planning, freedom of purchase
children’s clothing, treatment autonomy for children, treatment autonomy for own,
freedom for expenditure, knowledge about inheritance law, freedom of expenditure for
own, freedom of saving money, freedom of purchasing ornaments, free to move outside,
opinion for land dispute, freedom to travel, freedom to purchase of properties for own,
whether microcredit holder. Using these variables, a composite level of women’s
empowerment was assessed by Principle Component Analysis. After getting the score
they categorized empowerment as poor with score 1-82, fair with score 83-122 and good
with score above 123. The authors showed that 40% had poor, 19% had fair and 41%
had good level of women’s empowerment. Using bivariate analysis this study has
revealed that level of women’s empowerment is high among the respondents living in
the urban areas, having less number of children, religion in Islam, having higher level of
education of both husband and wife, engaged in jobs, having higher monthly income
contributed to family and living in pacca house (p<0.05). It also shows that there are no
72
significant association with current age of both husband and wife, age at marriage and
spousal age difference (p>0.05). This study does not clarify how the scores were given to
assess empowerment and it considers only individual factors at household level.
Reviewing the studies related to women’s empowerment we can easily select the major
dimensions and their indicators of the women’s empowerment those help us quantify the
women’s empowerment. We are also enlightened with the several socio economic and
demographic outcomes of women’s empowerment. Besides, the detailed review of
literature helps us recognise the important determinants of women’s empowerment in the
district of Bankura, West Bengal.
2.5 Conclusion
The literature review in this chapter has covered the wide range of vantage points for
investigating women‘s empowerment like, autonomy (Dyson & Moore 1983; Kabeer,
2001; Jeejebhoy & Sathar 2001), agency and status (Jain et al. 2003, Kumar, 2011),
women‘s land rights (Quisumbing et al., 1999), process of gaining control over self and
resources (Verma, 2009, Samanta, 2009), domestic economic power (Handy, et al.
2004), bargaining power (Beegle et al., 1998; Quisumbing & de la Briere 2000), power
(Agarwal, 1997; Beegle et al., 1998), patriarchy (Malhotra et al., 1997), gender equality
(World Bank, 2001 & 2012). Often there is not any clear demarcation in the meanings of
these terms for women’s empowerment. A few studies have attempted to develop the
index for women’s empowerment. Many studies examined the impact of different socio-
economic-demographic factors on women’s empowerment. A wing of studies has tried
to examine the impact of empowerment on several dimensions of family and child
welfare. In this dissertation applying a sophisticated econometric tool we develop the
index for women’s empowerment at the individual level as well as community level.
Then we examine the impact of empowerment on three important aspects of family and
child welfare and explore the important determinants of women’s empowerment in
Bankura district. This type of thorough study on women’s empowerment is still
uncommon in existing literature. In the coming chapter we come to the details of models,
methodology and data source of our study.
73
Chapter Three __________________________________________________
Model, METHODOLOGY AND DATA
3.1. Introduction
In the second chapter we have reviewed the literature on the studies related to women’s
empowerment. In the course of the literature review, we have seen that many studies
deal only with the concept of women’s empowerment. Majority of the theoretical studies
suggest that women’s empowerment is multidimensional and context specific. A number
of studies have tried to quantify the qualitative idea, empowerment, particularly
women’s empowerment. Some of the empirical studies have tried to find out the
important determinants of women’s empowerment. The impacts of women’s
empowerment on human well-being have been assessed in a few studies. Different
studies have taken different indicators or measure of women’s empowerment in
accordance with their context and objectives. They have applied different methodologies
for estimating the impact of women’s empowerment on household and child welfare. In
order to capture the concrete idea of women’s empowerment in Bankura District of West
Bengal we specify the suitable econometric models and methodologies in this chapter.
We have planned to divide our study in three parts. In the first part of this study we want
to measure the level of empowerment for the rural women in the district of Bankura.
Second, we are interested to find out impact of empowerment on the household welfare
as well as on human wellbeing. In this part we have planned to emphasize on the three
issues of household welfare, viz. issue of family planning, issue of domestic violence
and issue of child education. In the third part of our dissertation the various influential
factors of women’s empowerment in the district of Bankura are dealt with. In order to
carry out our study with the issues mentioned above we need to find out a suitable
74
approach to quantify empowerment and relevant methodology for estimating the
different models related to the issues of our study. With this end in view, we have
arranged the current chapter in the following manner.
In section 3.2 we have developed two systematic approaches for quantifying women’s
empowerment along with its different dimensions. Sub-section 3.2.1 deals with the
selected dimensions and corresponding indicators of women’s empowerment at the
individual/household level as well as at the community level. We explain the details for
measuring the degree of empowerment in sub-section 3.2.2. In sub-section 3.2.3 we have
explained the framework for developing a composite index for women’s empowerment.
This section will help us show the empowerment status of women in the district of
Bankura. Section 3.3 deals with the research design for studying the impact of women’s
empowerment on household welfare. This section has been decomposed into three sub-
sections. In sub-section 3.3.1 we have explained the methodology for studying the
impact of women’s empowerment on family planning. We have addressed the issue of
domestic violence which is expected to be connected with empowerment. Analytical
framework for studying the nature of domestic violence against women in the district of
Bankura has been presented in sub-section 3.3.2. Sub-section 3.3.3 has presented the
analytical framework for the study of empowerment on child education. The analytical
framework for finding out different influential factors of women’s empowerment has
been explained in section 3.4. Regression specifications for each of the selected issues of
women’s empowerment have been presented in section 3.5 along with its subsections.
The definitions and measurements of the dependent and independent variables of the
specified models specified have been explained in Section 3.6. In section 3.7 we have
logically developed the hypotheses relating to the specific econometric models. The
methodology of data collection and the diagnostic check for the sample size have been
discussed in Section 3.8 and in its different sub-sections. Finally, we have come to the
conclusion of this chapter in section 3.9. Now we are going to present the detailed
analytical frameworks one by one.
3.2. Measures of Women’s Empowerment
In this section we want to quantify the idea of women’s empowerment. We have already
discussed that women’s empowerment is a buzzword now-a-days in developmental
policies. It is qualitative and multidimensional in nature. However, there is not any
75
universally accepted definition of women’s empowerment in the existing literature; its
definition is very much context specific. One may consider women’s empowerment in
the individual/household context or in the community/locality context or in the
national/international context. Consequently different studies have tried to measure it in
different ways depending on different contexts. Our study area, the district of Bankura, is
mainly a rural as well as poor district in the state of West Bengal. In this district the
major development indices of women are standing below the expected level. Due to the
geographical as well as social backwardness it is not surprising that women of this
district have very little power and voice in the broader area of life. In other words, the
empowerment of women in the national or international level in the area under study is
almost zero and invariant in the broader area of life. However, owing to the different
local and social customs of different communities, women of different regions and
religions and communities have enjoyed different levels of empowerment at the
household level as well as at the community level. Against this situation if one attempts
to focus empowerment from a single window, it will give the wrong impression about
the actual empowerment of a particular woman. In order to address the possibility of this
wrong impression in the measure of empowerment we have sketched to study
empowerment at the household level and at the community level separately. Therefore,
we would like to measure empowerment of women in two levels.
Women’s Empowerment at the Household Level
Women’s Empowerment at the Community Level
In the course of literature review we have found two wings of studies of quantifying
women’s empowerment. One wing has considered the value of a specific indicator or
simple average of the values of the selected indicators/dimensions of empowerment
(Ghuman, et al. 2004, Srivedi, 2005, Adhikary and Dutta, 2011, Varghese, 2011). Other
wing has tried to cover a wide range of the indicators and considered the weighted
average of the values of the indicators of empowerment as a measure (Kishor, 1997,
Jejeebhoy, 2000, Koenig, et al. 2003, Handy et al. 2004, Parveen et al. 2004, Chowdhury
et al. 2009). However, no one of the studies takes into account both the types
simultaneously and compares the pros and cons of these two types of measures. With
this end in view, we have planned to measure women’s empowerment at household as
well as at community levels by two different ways – simple average method and
76
weighted average method. In this study Principal Component Analysis has been used to
derive the weights of the indicators of women’s empowerment. Before going to the
detailed specification of the methodology for measuring women’s empowerment we
discuss the dimensions and selected indicators of women’s empowerment at the
household level and at the community level.
3.2.1. Selected Dimensions and Indicators of Women’s Empowerment
We have found several dimensions and indicators of women’s empowerment in the
existing literature. A section of existing literature only proposed the dimensions and
indicators for measuring empowerment. Different studies reviewed in chapter two have
tried to measure empowerment empirically considering different dimensions and
indictors which are fitted in their context. In an exhaustive analytical study, Molhotra et
al. (2002) has proposed several indicators for measuring empowerment at household
level, at community level and at broader arenas. She has also proposed six dimensions of
women’s empowerment for each level. These are economic, socio-cultural,
familial/interpersonal, legal, political, and psychological. However, she did not quantify
women’s empowerment in practice. Actually these dimensions are extremely broad in
scope and not very much easy to capture at a time. In our dissertation we have planned to
quantify women’s empowerment at the household level as well as at the community
level. We have tried to follow the dimensions as proposed by Molhotra, et al. (2002). In
order to cover the dimensions we have selected some relevant criteria as indicators in the
context of our study. The selected indicators for respective dimensions are as follows.
3.2.1A. Dimensions and Indicators of Women’s Empowerment at Individual or
Household level
Indicators of Economic Dimension
Whether she has control over her personal income or asset.
Whether she can have access to household resources.
Proportion of household expenditure that she bears. (>50% | < 50% | None)
Who decide the use of saving/ loan? (Own / with spouse/ with other family
member/other members)
Does she take part in the decision for selling or buying asset for household?
Does she enjoy freedom in choosing her occupation? (Yes/No)
77
Indicators of Political Dimension
Did she cast her vote in the last election?
Does other influence her to cast her vote?
Does she know the name of local leader? (panchayat pradhan / councilor/ MP/
MLA)
Whether she know the candidate of opposition party in the last election.
Does she get domestic support for her political engagement?
Indicators of Socio-Cultural Dimension
Whether she is free to move outside her home.
Does she regularly enjoy Radio, telephone, TV and Newspaper?
Whether or not she participates in local cultural programs.
Does she want to educate her girl or other girls in her household?
Does she arrange the marriage of the girls before their eighteen years old or
support it?
Whether she want to send her child for earning.
Indicators of Personal / Familial Dimension
Whether her marriage is arranged or self selection.
Can she articulate her personal problem to other family members?
Whether or not she can independently decide about her child education, health,
food etc.
Whether she has decision making power regarding her personal health, body.
Whether anybody interfere when she talks to strangers.
Indicators of Legal Dimension
Whether she knows the mechanisms of justice used in the locality.
Does she think women/men get (better, equal or worse) treatment from this
system?
Whether she knows the laws and legislation available in favour of women.
Whether she knows about the various kinds of public services available in the
locality.
Whether her marriage is registered or not.
78
3.2.1B.Dimensions and Indicators of Women’s Empowerment at Community level
Indicators of Economic Dimension
Whether she is employed/earner or not.
Whether she has ownership of land or property or not.
Whether she has access to formal savings, insurance or loan etc.
Whether she has access education or training service when she needs it.
Whether anybody threats her to evict from property.
Whether her present occupation is secured or not.
Indicators of Political Dimension
How much does she involve in political process? (Very involved / slightly
involved / not at all)
Whether she attends any political gathering or not.
Whether she is a member of any political party.
Did she ever contest vote as a representative?
Is she leader of any organization?
Indicators of Socio-Cultural Dimension
Does she participate in community activity?
Whether she is a member of any social organization or group.
Whether she can influence the election/ selection of the leader of organization or
group.
Whether or not she knows the location of the nearest post-office, school, hospital,
club, vegetable market, other social/ cultural organization.
Does she feel exclusion from participation in any community activity organized
by local government, religious organization, school, the local development
association etc.?
Does she oppose the social curses like a) Dowry system, b) Inter-caste marriage,
c) preference of male child?
Indicators of Personal / Familial Dimension
Did she ever campaign against social curse like dowry, violence?
Whether she has professional training or not.
79
Has she voluntarily changed her occupation after marriage?
Did she sacrifice employment or membership of any organization due to familial
ground?
Whether she has immunised her children in due time.
Indicators of Legal Dimension
Whether she ever used the mechanism to seek redress or access justices.
Whether she complains about the deficiency of public services in her locality.
Is she active in complaining about any problem to the system of justice? (Very
active, a little bit active, fairly active, and not active at all)
Does she think that authorities are more, less or equal effective about other
people’s need/concern compared to her? - (More equal, less).
3.2.2. Degree of Women’s Empowerment
Let us now quantify women’s empowerment based on the above indicators. We intend to
measure empowerment index by two methods. First, we would like to compute the
degree of women’s empowerment for each woman at household level and at community
level using simple average method. Therefore, the simple indices adopted in the study
for measuring the empowerment of women at the household level and at the community
level are given as follows.
100level household at the criteria selected ofnumber total
womanby the fulfilled level household at the criteria selected ofnumber DOWEH
100level community at the criteria selected ofnumber total
womanby the fulfilled levelcommunity at the criteria selected ofnumber DOWEC
where, iDOWEH and iDOWEC are the degree of women’s empowerment at the
household level and at the community level for ith woman.
Using this formula we can easily measure the empowerment of women at the household
level and at the community level, which are now quantitative in nature. In accordance
with the specification of the quantification of the degree of empowerment we can say
80
that the values of women’s empowerment vary from zero to hundred. The higher the
value of empowerment index the higher is the empowerment of women i.e. the higher is
the status of women in household and in society.
3.2.3. Analytical Framework for Computing Composite Index of Women’s
Empowerment
In order to study the empowerment level of the women at the household level and at the
community level we have considered a large number of indicators which are usually
correlated because they are measuring the same issue. So to make the analysis concrete
we need to reduce the large number of indicators to lesser number of factors that are
being measured by the indicators. Besides, we have put same weight for each
dimension/indicator in our first measure; but in reality different dimension/indicator
should have different weight in contributing to women’s empowerment. The weights of
the indicators may be derived in a number of ways. Factor analysis is the common
statistical tool for the derivation of relative weights and is applied to deduce a set of
latent factors that account for the patterns of collinearity among multiple metric
variables. In literature we have found a variety of methods that extract factors from an
inter-correlation matrix of the variables. Principal Component Analysis (PCA) is
probably the most common method used in the extraction of factors. In order to study
and understand the level of empowerment extracting a small number of
orthogonal/uncorrelated variables called Principal Components from the set of indicators
under consideration and to derive relative weights of the selected indicators we would
like to use the technique of Principal Component Analysis. Each Principal Component is
a linear combination of all the variables/indicators under studied having unit variance.
In PCA, the linear model accounts for large proportion of the variation of the data set. If
the PCA becomes unrotated, the eigenvectors may not align close to the data clusters and
thus may not represent the actual physical states as well. The rotated PCA methods rotate
the PC eigenvectors so that they become closer to the cluster of data point. There are
several strategies for rotation in the literature of Econometrics. Orthogonal rotation
strategy is applied if no relationship is found between the principal components. In order
to determine whether the extracted components are related or not, a simple correlation
has been applied on the component scores. If the results show no correlation one can
apply orthogonal rotation strategy on the indicators. In the present study Varimax
81
rotation strategy, which is the most popular form of orthogonal rotation schemes, has
been applied. Varimax rotation maximizes the variance of the rotated squared PCs. It
means that the squared loadings are made as large as possible or as close to zero as
possible means that many of the loadings are essentially set to zero, yielding loading
patterns which have more localized features than the unrotated patterns. In view of the
above, PCA has been re-run/rotated (Varimax) specifying the fixed number of
components (whose Eigen value is greater than one) to be retained. The respective
rotated component scores have been obtained by regression method. In order to develop
a composite index of women’s empowerment we derive the component scores also
(Antony et al. 2007). The weighted sum of the component scores may be considered as
the composite index of women’s empowerment.
Finally, in order to arrive at the composite index of women’s empowerment (CIWE), the
rotated component scores and the corresponding per cent of variances (rotation sums of
squared loadings) accounted by the principal components are used. Principal
Components are extracted in decreasing order of their variances. This indicates that
variances explained different principal component are different, thus the weight of
different principal components are different. To address this, we have constructed the
CIWE as the weighted sum of the component scores – the weights being percentage of
variations explained by the respective Principal Components after rotation. Thus, CIWE
is to be calculated taking the sum of product of Component score ( ikC ) and
corresponding percent of variance kV explained by the principal component (Antony et
al, 2007)
ikki CVCIWE (3.2.1)
where, iCIWE stands for the composite index of empowerment for ith woman, kV
denotes the percentage of variation explained by kth principal component after rotation
and ikC is the component score of ith woman with respect to kth
principal component
after rotation.
Applying PCA tool for the indicators of women’s empowerment at the household level
we derive this formula and compute the CIWEH for the sample women at the household
82
level. Similarly for the indicators of women’s empowerment at the community level we
compute the CIWEC for the sample women at the community level. It should be noted
that the values of these indices vary from to and follows standard normal
distribution. So it has some good properties for inference analysis.
3.3. Research Design for Studying the Impact of Women’s empowerment on
Household Welfare
In the new millennium women’s empowerment has been recognized as an instrument for
improving household welfare. A large section of the literature regarding women’s
empowerment has put their attention on quantification and on measurement of women’s
empowerment from several view points. A group of studies has tried to identify the
responsible factors for empowering women. A very few studied have examined the
impact of women’s empowerment on several dimension of household welfare. Again
among these studies a section has theoretically analyzed the expected gain of women’s
empowerment on household welfare. Systematic empirical estimation of the impact of
women’s empowerment on household welfare is very much limited. Against this
situation, we are interested to estimate the impact of women’s empowerment on three
specific issues of household welfare. These three issues are the family planning,
domestic violence and child education expenditure. In the subsequent sub-sections we
deal with the research design for studying the impact of women’s empowerment along
with some other selected variables on decision towards family planning, on incidence of
domestic violence and on child education expenditure in the context to our sample
women in Bankura district.
3.3.1. Analytical Framework for Studying Impact of Women’s Empowerment on
Family Planning Decision
We like to study how the decision regarding family planning is affected by women’s
empowerment. In order to estimate the impact of women’s empowerment along with
other explanatory variables on the family planning decision we formulate probit model.
The decision regarding family planning is the dependent variable in this particular case.
The decision towards family planning is specified by the information whether or not the
woman is an ideal family planner. If the woman has only one or two children and there is
no possibility of further child adaptation, or if the woman has strong decision that she
will not have more than two children, the woman is identified as an ideal family planner.
83
The couples who have more than two children or have not taken any decision about
family planning are considered as non-family planner. In this view, each woman has
only two possible responses, ‘yes’ and ‘no’ regarding family planning decision.
Therefore, the variable, decision regarding family planning can take only two values for
two responses. We attach value ‘1’ for the woman who has taken family planning
decision and ‘0’ otherwise. The decision regarding family planning ( iY ), which is the
dependent variable, is a dichotomous variable taking values ‘1’ and ‘0’.
The family planning decision definitely depends on some socio-economic and
demographic traits of the woman and her household and on some community traits. It is
expected that empowerment of the particular women have a crucial role for taking family
planning decision. This means that women’s empowerment along with several socio-
economic and demographic traits of the woman/household determine the decision
towards family planning. Let iP stands for the conditional probability of taking decision
regarding family planning. In the present framework, in terms of a typical binary
response model we like to assess iP conditional on certain information set, , which
consists of socio-economic and demographic traits of individual/household along with
women’s empowerment at the household level and at the community level. These traits
are apparently exogenous and predetermined variables that have been considered as
determinants of the conditional probability of taking family planning. Already we have
expressed iY , as dichotomous variable and iP is the probability that 1iY Therefore, by
definition of mathematical expectation we can say iii XYEP , where iX stands
for socio-economic and demographic traits of the thi individual/household along with
women’s empowerment belonging to the information set, . Hence, the decision
regarding family planning for the sample women is a function of the information set, .
Factors affecting the Decision towards Family Planning
We have classified the socio-economic and demographic traits affecting the decision
regarding family planning into two categories, namely, the individual and household
characteristics and the community characteristics of the woman. Examination of the
impact of women’s empowerment at the household level and at the community level on
the family planning decision is one of our prime objectives. So, women’s empowerment
84
at the household level is the main explanatory variable under the individual/household
characteristics. In addition to women’s empowerment at the household level, we
consider some other variables as determinants of the decision regarding family planning.
In the category of individual and household characteristics we have incorporated male
child bias, education level, age at marriage, spousal age gap, spousal education, family
composition, household’s land holding, household occupation, per capita family income,
dependency ratio. We expect that women’s empowerment at the household level should
have a positive impact on the decision regarding family planning for the sample women.
In order to check the empirical validity of the claim we include women’s empowerment
at the household level and at the community level as chief exogenous variables in the
model for the decision regarding family planning. In the rural society we see that
education level of the woman as well as other family members affect the outlook to
family planning. Age at the time of marriage of the woman is crucial for taking any
fruitful decision regarding family planning. Dependency ratio is another determinant of
the family planning decision. Besides, we cannot ignore the variables like household’s
landholding, per capita family income and occupational status of the family as important
determinants of the decision regarding family planning.
There are several community characteristics like caste, race, religion, region, social
customs that affect the family planning decision of the couples. But in our dissertation
we have planned to study the attitudes of rural couples regarding family planning in the
district of Bankura. In the area under study the variation of region, race, religion, social
custom are very much limited. With this end in view, women’ empowerment at the
community level, participation in SHG-centric microfinance program and caste, three
community characteristics, have been considered as important exogenous variables
affecting the decision regarding family planning in this study. In order to analyze the
impact of caste we have categorized the persons in four groups like Scheduled Caste,
Scheduled Tribe, Other Backward Classes and General Caste as per government
regulation. Against this backdrop, we can write
Decision regarding Family Planning= f (Women’s Empowerment at the household
level and at the community level, Individual/Household Characteristics, Community
Characteristics, and Random Disturbances)
85
On the basis of such specification, we finally pay our attention in Probit Model. The
cumulative distribution function for Probit model is given by
dx22x
1exp
2Π
1F(x)
(3.3.1)
where x is a standard normal variable and xF is a cumulative distribution function.
3.3.2 Analytical Framework for Impact of Empowerment on Domestic Violence
against Women
In order to assess the impact of women’s empowerment on household welfare, we would
like to examine the impact of women’s empowerment at the household level and at the
community level along with other determinants on the probability of the incidence of
domestic violence that she faces. The incidence of domestic violence is
multidimensional in nature. So, to study the extent of domestic violence we need to
specify the several indicators for measuring the intensity of domestic violence. But in the
course of our field survey we have observed that most of the rural women are not willing
to disclose the intensity of violence from that they suffer. That is why; we are reluctant
to form a composite index for domestic violence for our sample women. We have
collected information from the woman or from her neighbour whether or not the
particular woman faces frequent physical violence from her husband or from other
household members. If a woman suffered from physical assault by her husband or other
family member at least in two episodes during the last six months, we have identified
that the particular woman faces domestic violence. With reference to our criteria,
therefore, domestic violence is a binary variable. Let us now discuss the several socio-
economic-demographic and community traits that may affect domestic violence against
women.
Factors Affecting Domestic Violence against Women
We can divide the determinants of domestic violence against women into two categories,
namely, individual/household characteristics and community characteristics. In the
course of our literature review we have noticed that individual/household’s socio-
economic-political factors such as nature of marriage, duration of marriage, dowry,
women’s education, age, occupation of woman, childlessness, economic pressure, and
86
intergenerational transmission of violence, financial condition of family, husband’s
education, average education level of the family, family type, household occupation are
important determinants of the incidence of domestic violence against women. On the
other hand, different studies have reported that caste, region, religion, environment of
violent crime and community-level norms concerning wife beating, culture and customs
of the society are the major community level factors that influence the incidence of
domestic violence against women.
In addition to women’s empowerment at the household level and at the community level
in the spectrum of individual/household characteristics we have taken duration of
married life, spousal age gap, family type, highest male education, husband education,
household occupation, husband’s drug addiction, annual per capita income, land holding,
dependency ratio, dowry demand at and post marriage as determinants of the incidence
of domestic violence against women in the category of individual/household
characteristics. Among the community characteristics we have considered women’s
empowerment at the community level, duration of self-help group membership and
caste. Note that caste of the sample member has been divided into four categories as we
have mentioned in the last sub-section. Hence, we can write:
Domestic Violence against Women = f (Women’s Empowerment at the household level
and at the community level, Individual/Household Characteristics, Community
Characteristics, Random Disturbances)
The incidence of domestic violence is a dichotomous variable (iY ) that takes the value
‘1’ for the woman who is victims of domestic violence and value ‘0’ for others. So, in
order to estimate the probability of facing domestic violence for a woman in the district
of Bankura we formulate a Logit model. The logistic function for the logit model is as
follows
)exp(1
)exp()exp(1)(
1
X
XXX
(3.3.2)
where )(X expresses the cumulative distribution function of X which follows logistic
distribution.
87
3.3.3. Analytical Framework for Impact of Women’s Empowerment on Children’s
Educational Expenditure
The third issue that we have selected for studying the impact of women’s empowerment
on household welfare is the nature of educational expenditure for children. It is true that
the women/households, who have no children in the age of studying in educational
institutions, don’t have this expenditure at all. On the other hand, the women/households,
who have children in the age of studying in educational institutions, have two options:
either send their children school/colleges for education or not. The women/households
who send their children for education must have some expenditure towards children
education. But expenditure for children education is also zero for the women/households
who do not send the children in educational institutions. Actually, this part of our study
is restricted among those sample women who have children in the age of attending
school and colleges. A very large portion of our sample women have children in the age
of attending school and colleges. So this part of our study does not lose generality. The
educational expenditure for children in a family is one of the major indicators of a family
wellbeing. It is expected that development of the family increases the expenditure for
children’s education. This study has planned to investigate the impact of women’s
empowerment at the household level and at the community level on education
expenditure for her children. Education, particularly, primary education is most
important for economic development. In India, as well as in our state of West Bengal,
primary education is the four years of schooling from the age of six. The duration of
education up to the class of eighth standard is termed as elementary education. Education
up to tenth class is known as secondary education. Education after secondary level is
viewed as higher education. Education up to elementary level has been made compulsory
and free for all the children by the government of India. But we have the experience that
a large section of the guardians of the students at any level cannot depend completely on
the government and government aided schools for education of their children. Most of
them arrange private tuition for their children. In urban area a section of guardians send
their children in private school. However, in rural area, like our study area this
opportunity is completely absent. A major part of the guardians, who like to educate
their children, in rural area also send their children to private tutors. However, during the
course of field survey we have observed that most of the people in our study area are
living in a poor economic condition. A considerable number of children have to engage
in job to collect their livelihood. In our study area many potential students could not
88
complete their primary education. Here dropout rate is high. As a result, educational
expenditure varies widely across the sample households. Like any other part of India
here primary and secondary education are provided by the government, forming a free
education system. In spite of these, a major number of households have to spend money
for education of the children.
The expenditure relating to education of the children is viewed as children’s educational
expenditure. Expenditure for purchasing educational kits like books, papers, pens,
pencils school dresses etc., are the essential expenditure for children’s education. The
fact is that most of the students at primary level as well as secondary level even at the
college level take the shelter of private tuition. The major part of expenditure for
children’s education is the fees for private tuition. Almost all villages in our country
have at least one primary school. So traveling cost for attending primary school is zero
for almost all households. However, women/households have to bear some travelling
cost for their children who attend secondary or higher secondary school and college.
This cost proportionally amplifies with the number of the tuition trips. Therefore, cost
for commuting school/colleges and tuition is a significant part of children’s educational
expenditure. Another head of educational expenditure is for school (private school) and
college tuition fees. The students, who study at residential institution or study in the
institution far away from their residential addresses, need some accommodation cost.
Besides, women/households give some money to their children for refreshment during
travelling for education. In addition to these, actual costs for children’s education there
may have some opportunity costs. It is common that during the time of education parents
or other family members have to spend some times every day for preparing and or
escorting the children for school and for coaching. Sometimes guardians have to attend
the school/colleges for admission and examination of the children, meetings or for
functions. These factors give rise to some opportunity cost relating to children’s
education. Other part of the opportunity costs of children’s education arise when the
children have some opportunity to earn during the period of education. Sometimes,
guardians want to send their child for earning. This fact also adds some opportunity cost
for children’s education to the guardians. In order to explain the analytical framework
for modeling the child education expenditure first note down the probable heads of the
educational expenditure for children.
89
Probable Heads of the Educational Expenditure for Children
Expenditure for purchasing educational kits including school uniform if any
The cost for private tuition fees
Commuting expenditure for attending tuition and school
Fees for education (for higher education/private school)
Accommodation costs if any
Tiffin cost/ pocket money
Opportunity cost ( for the guardians themselves and for the guardians who wants
to send their child for earnings)
We did not consider the opportunity cost for children’s education due to lack of reliable
information and calculating hazards. In estimation we have considered annual child
education expenditure as proportion to annual household income. To compute
expenditure for children’s education first we calculate the expenditure for each head for
month or for year as we have been informed from the respondent and add them. Then we
average the expenditure for the last year. We have finally divided the annual child
education expenditure by annual household income to get the proportion of household
income spend for child education. It is a quantitative variable and expressed as
percentage.
Factors affecting Children’s Educational Expenditure
No doubt the educational expenditure varies from household to households depending on
the number children and their educational level. It is also directly related with income of
the households. In addition to these two factors, children’s educational expenditure
depends on mother’s/household characteristics and community characteristics.
Educational expenditure depends on certain information set, which consists of socio-
economic and demographic traits of women along with her empowerment status. These
traits are exogenous and predetermined variables that have been considered as
determinants of the educational expenditure for children.
We have classified the socio-economic and demographic traits affecting the child
education expenditure into two categories, namely, the individual/household
characteristics and the community characteristics of the woman. As per our specific
90
objective in the range of individual/household characteristics we have considered
women’s/mother’s empowerment at the household level as the primary explanatory
variable determining children’s educational expenditure. Generally, empowered woman
at the household level as well as the community level are more conscious about the
effectiveness of education. Against this backdrop, we can write
Children’s Education Expenditure as Proportion to Household Income = f (Mother’s
Empowerment at the household level and at the community level, Mother’s/Household
Characteristics, Community Characteristics, Random Disturbances).
In order to estimate the impact empowerment status along with other explanatory
variables on the children educational expenditure as proportion to household income we
formulate a multiple log linear regression model as
iii UβXY ln (3.3.3)
where iY stands for the educational expenditure for children, iX denotes the set of
explanatory variables, which includes all the individual/household and community
characteristics along with the empowerment variables, k captures the marginal impact
of the particular explanatory variable on (ln iY ) child educational expenditure as
proportion to household income and iU are the random disturbances.
The estimation of this model will help up capture the rate of change of the expenditure
for children’s education for change in empowerment at the household level and at the
community level.
3.4. Analytical Framework for Studying Women’s Empowerment
We expect that women’s empowerment is an instrument for improving household and
child welfare. Particularly we have already formulated the framework for assessing the
impact of women’s empowerment on family planning decision, on incidence of domestic
violence against women and on child education expenditure. In this connection it is
rational to quest the factors affecting household level and community level of
empowerment. With this end in view, in the present section, an attempt has been taken to
describe the strategy to estimate the women’s empowerment at the household level and
91
at the community level so that we can find out the probable determinants of it in the
context of our study objectives and study area.
Factors affecting Women’s Empowerment
Several studies reviewed in chapter two have examined the impact of several socio-
economic and demographic traits on women’s empowerment. Similarly in the context of
our study we have selected some factors affecting women’s empowerment at the
household level and at the community level. In the range of individual/household
characteristics we have considered age, education, personal occupation and income, per
capita household income, household occupation, landholding, family type, educational
background of the household and access to formal credit. We have considered SHG
membership and caste in the range of community characteristics.
Therefore, the general functions for studying empowerment at the household level and at
the community level can be written as follows.
Women’s Empowerment at the Household Level = f (Individual/Household
Characteristics, Community Characteristics, Random Disturbances).
Women’s Empowerment at the Community Level = f (Individual/Household
Characteristics, Community Characteristics, Random Disturbances).
In order to estimate the empowerment status as a function of some individual/household
characteristics and community characteristics we have planned to formulate multiple
linear regression models. In this analytical framework the indices of women’s
empowerment are the dependent variables which are quantitative by construction. Now,
we can write the function for each index of women’s empowerment in terms of general
linear regression model as
iii UβXY (3.4.1)
where iY stands for the index of women’s empowerment of the ith
woman, iX denotes
the set of explanatory variables, which includes all the individual/household and
community characteristics affecting the empowerment status, k captures the marginal
92
impact of the particular explanatory variable on empowerment index and iU are random
disturbances.
3.5. Regression Specification of the Analytical Models relating to Women’s
Empowerment
Once we have the theoretical framework of the models corresponding to the issues of the
women’s empowerment, we need to specify the empirical models for estimation. Now
we specify the econometric models serially in accordance with the analytical
frameworks.
3.5.1. Probit Models for Decision regarding Family Planning
In order to specify the probit model for the decision regarding family planning we follow
the framework analyzed in section 3.3.1. In this context we have considered two models,
Model-1A and Model-1B. In Model-1A we have taken the degrees of women’s
empowerment as explanatory variables and in Model-1B the composite indices of
women’s empowerment have been considered as explanatory variable along with some
other explanatory variables. The specific models are as follows.
Model-1A Probit Model when Empowerments Quantified by Simple Average
USTSCOBCDSHGMDOWECAPCHINDRATIO
HLANDNONFARMCULTITYFAMIHEDUEDU
SAGEGAGAMFFSFFFSMFMSFDOWEHDRFP
19181716151413
121110987
6543210
(3.5.1)
Model-1B Probit Model when Empowerments Indexed by PCA
USTSCOBCDSHGMDOWECAPCHINDRATIO
HLANDNONFARMCULTITYFAMIHEDUEDU
SAGEGAGAMFFSFFFSMFMSFDOWEHDRFP
C
C
19181716151413
121110987
6543210
(3.5.2)
93
3.5.2. Logit Models for Incidence of Domestic Violence
As the incidence of domestic violence against women is a dummy variable in accordance
with our specification, we formulate a Logit model. We have measured women’s
empowerment levels following two methodologies. Keeping this view in mind, here we
also estimate two models. Degrees of women’s empowerment have been considered as
determinant of domestic violence against women in Model-2A whereas Model-2B
includes composite indices of women’s empowerment.
Model-2A Logit Model when Empowerments Quantified by Simple Average
USTSCOBCDSHGMDOWECADDICPMDOW
DOWAPCHINDRATIOHLANDNONFARMCULTI
TYFAMIHIMEDUHEDUSAGEGDURMDOWEHDVIO
19181716151413
121110987
6543210
(3.5.3)
Model-2B Logit Model when Empowerments Indexed by PCA
USTSCOBCDSHGMCIWECADDICPMDOW
DOWAPCHINDRATIOHLANDNONFARMCULTI
TYFAMIHIMEDUHEDUSAGEGDURMCIWEHDVIO
C
C
19181716151413
121110987
6543210
(3.5.4)
3.5.3. Log-Lin Models for Children’s Education Expenditure as Proportion to
Household Income
In the section of analytical framework we have explained the justification of log-lin
model for examining the impact of women’s/mother’s empowerment on household’s
expenditure for children’s education. In this sub-section we express the explicit form of
the regression models. Like another issues in this case we also set two models - one is
based on the degrees of women’s empowerment and other on the composite indices of
women’s empowerment.
94
Model-3A Log-Lin Model when Empowerments Indexed by Simple Average
USTSCOBCDSHGMDOWEC
APCHINLOGDRATIOHLANDNONFARMCULTI
TYFAMIHIFEDUHIMEDUHEDUDOWEHLEDEX
1514131211
109876
543210
)(
(3.5.5)
Model-3B Log-Lin Model when Empowerments Indexed by PCA
USTSCOBCDSHGMCIWEC
APCHINLOGDRATIOHLANDNONFARMCULTI
TYFAMIHIFEDUHIMEDUHEDUCIWEHLEDEX
C
C
1514131211
109876
543210
)(
(3.5.6)
3.5.4. Linear Regression Models for Women’s Empowerment at the Household
Level and at the Community Level
By definition and conception the empowerment of women is a qualitative attribute of the
women. So far we have quantified it at two levels, namely at individual/household level
and at community level. Further, to measure the women’s empowerment for each level
we have considered two methodologies. Applying simple average method we have
computed degrees of women’s empowerment at the individual/household level and
women’s empowerment at the community level. We have also computed composite
index of women’s empowerment at the individual/household level and at the community
level applying Principal Component Analysis. In the previous section we have explained
the analytical framework for studying women’s empowerment at household level and
community level. Therefore, we have four measures for women’s empowerment. To this
end in view, we formulate four linear regression models to identify the major
determinants of women’s empowerment in the district of Bankura. First two are related
with household level empowerment and last two with community level empowerment.
These have been specified as follows.
95
Model-4A Linear Model for Household Level Empowerment Indexed by Simple
Average
USTSCOBCDSHGMHIFEDUHIMEDU
HLANDNONFARMCULTIAPCHINDRATIOTYFAMI
AFCTPINCSELFLABEDUAGEAGEAGEDOWEH
201918171615
14131211109
876543210 321
(3.5.7)
Model-4B Linear Model for Household Level Empowerment Indexed by Principal
Component Analysis
USTSCOBCDSHGMHIFEDUHIMEDU
HLANDNONFARMCULTIAPCHINDRATIOTYFAMI
AFCTPINCSELFLABEDUAGEAGEAGECIWEH
201918171615
14131211109
876543210 321
(3.5.8)
Model-4C Linear Model for Community Level Empowerment Indexed by Simple
Average
USTSCOBCDSHGMHIFEDUHIMEDU
HLANDNONFARMCULTIAPCHINDRATIOTYFAMI
AFCTPINCSELFLABEDUAGEAGEAGEDOWEC
201918171615
14131211109
876543210 321
(3.5.9)
Model-4D Linear Model for Community Level Empowerment Indexed by Principal
Component Analysis
USTSCOBCDSHGMHIFEDUHIMEDU
HLANDNONFARMCULTIAPCHINDRATIOTYFAMI
AFCTPINCSELFLABEDUAGEAGEAGECIWEC
201918171615
14131211109
876543210 321
(3.5.10)
96
In order to estimate Model-1A and Model-1B, we apply Binary Probit Maximum
Likelihood method. Model-2A and Model-2B have been estimated by Binary Logit
Maximum Likelihood method. The remaining models would be estimated using
Ordinary Least Squares method.
3.6. Definition and Measurement of the Variables included in the Regression
Models
We have specified the empirical models in the last section. In this section we define and
specify the measurement of the variables that have been considered in these models. We
would like to define and measure the variables considered in the whole study serially.
Note that a few variables are common in more than one model. To avoid the repetition,
we only refer the variables those have already been defined and explained.
Variables in Models for Family Planning Decision
In this model, we look into the determinants of the family planning decision for the rural
women. The dependent variable in the model is the decision towards family planning,
which we explain first and then explanatory variables of this model.
DRFP (Decision regarding family planning by the woman under study): It is the
dependent variable in Model-1A and Model-1B. If the woman strongly inform that she
has or will have utmost two children, we consider her as ideal family planner. Therefore,
decision regarding family planning is a dichotomous variable such that
DRFP = 1, if the woman’s family planning decision is positive
0, otherwise
Women’s Empowerment Variables
In Model-1A and Model-1B the main explanatory variables are the women’s
empowerment variables. These are constructed variables. In this study we have measured
women’s empowerment at two levels- household level and community level. Detailed
methods of construction of women’s empowerment variables have been explained in
sub-sections 3.2.2 and 3.2.3. Four measures of women’s empowerment have been
specified below.
97
A] DOWEH (Degree of Women’s Empowerment at Household Level): It is simply a
ratio of the number of criteria fulfilled by the woman to the total number of criteria set
for women’s empowerment at the household level. It has been expressed as percentage.
So it ranges from zero to hundred.
B] DOWEC (Degree of Women’s Empowerment at Community Level): Like DOWEH, it
is a ratio of the number of criteria fulfilled by the woman to the total number of criteria
for women’s empowerment at the community level. It is also expressed as percentage
form and thereby ranges from zero to hundred.
C] CIWEH (Composite Index of Women’s Empowerment at Household Level): It is the
weighted sum of the component scores – the weights being percentage of variations
explained by the respective Principal Components after rotation. By principles of
Principal Component Analysis these composite Index of women’s empowerment are unit
free but the values of this index varies from to .
D] CIWEC (Composite Index of Women’s Empowerment at Community Level): It is the
weighted sum of the component scores – the weights being percentage of variations
explained by the respective Principal Components after rotation. It is also unit free but
the values of this index varies from to .
The first two measures have been considered as primary explanatory variables in Model-
1A and the last two are the main explanatory variables in Model-1B. In addition to the
empowerment variables we have considered some other explanatory variables which are
common in both the models for decision regarding family planning.
Male Child Bias: In order to measure the male child bias we have divided the women in
accordance with the sequence of their children up to second issue into five groups,
namely, a) women having one child only (ONEC), b) women having first and second
children male, (FMSM), c) women having first child male and second child female,
(FMSF), d) women having first child female and second child male, (FFSM) and e)
women having first and second children female, (FFSF). Now if we find women having
at least one female child are less likely to adopt family planning we say that they are
male child biased.
98
Thus, male child bias is a categorical variable and the women, who have only one child
or who have male child in first and second issue, have been considered as reference
category for analyzing the impact of other groups. Therefore, the included dummy
variables are as follows.
FMSF =1, for the woman having first child male and second child female
0, otherwise
FFSM = 1, for the woman having first child female and second children male
0, otherwise
FFSF = 1, for the woman having first and second children female
0, otherwise
AGAM (Age of the Woman at the Time of Marriage): It is the chronological age,
measured in years, of the woman at the date of her marriage.
SAGEG (Spousal Age Gap): Spouse Age Gap, measured in years, is the difference
between the physical age of husband and wife.
EDU (Educational Qualification of the Woman): It is the number of years the woman
attended the formal educational institutions.
HEDU (Educational Qualification of the Husband): It is the number of years the
husband attended the formal educational institutions.
TYFAMI (Type of Family): We consider two types of family – nuclear and joint.
Nuclear family is a family where the woman (wife) lives with her husband and their
children. On the other hand, joint family is one where husband, wife and their children
live with other members of family like mother-in-law, uncle, aunt etc. It is a dummy
variable as
TYFAMI = 1, if the woman belongs to nuclear family.
0, otherwise
99
Household Occupation: Household occupation means the economic activity in which
the workers of the family are involved in to earn their livelihood over years. We have
divided the households into three categories in terms of their main occupation as follows.
CULTI (Cultivation): Households mainly engaged in cultivation are considered in this
category.
NONFARM (Non-Farm Self-employment): If major portion of the family income comes
from the self-employed non-agricultural activities, the family is considered in this
category.
Wage Labour: Households, in which earning members are wage labours or earning from
the source other than cultivation or self-employment, are considered in this category.
In our study we have considered wage labour as the reference category for analyzing the
impact of other categories i.e., Cultivation (CULTI) and Non-farm Self-Employment
(NONFARM). Specifically,
CULTI =1, if the household is mainly engaged in cultivation.
0, otherwise
NONFARM = 1, if the household is mainly engaged in nonfarm self-employment.
0, otherwise
HLAND (Household’s Landholding): Household’s land is defined as the total size of
land owned by the household for cultivation or residence. We have taken landholding as
a quantitative variable, the unit of measurement of agricultural landholding being bigha
(1 bigha = 0.4 acre).
DRATIO (Dependency Ratio): Dependency ratio is defined as the proportion of
dependents or non-working members to the total members of the family. The
dependency ratio in the family is expressed as percentage.
APCHIN (Annual Per Capita Household Income): Household income is the sum total of
the incomes of all the earning members in the household. We have taken the total
monthly expenditure as proxy for the household’s monthly income and convert it into
100
annual income. Dividing it by household size we get per capita annual household
income. The unit of measurement of this variable is rupee.
Explanatory Variables Reflecting the Community Characteristics
The main community level variable in our study is women’s empowerment at the
community level. We have taken DOWEC and CIWEC as main explanatory variables
reflecting community characteristics in Model-1A and in Model-1B respectively. We
have already specified DOWEC and CIWEC. Now we define the other community level
variables affecting the decision regarding family planning, which are common in Model-
1A and in Model-1B.
DSHGM (Duration of SHG Membership): The duration of the SHG (self help group)
membership is measured by the period for which the woman acts as a member of the
SHG. We have taken month as measuring unit of the duration of SHG-membership.
Caste: Caste of a person is a categorical variable indicating the person belonging to a
specific caste, namely, General Castes (GEN), Other Backward Classes (OBC),
Scheduled Caste (SC) and Scheduled Tribe (ST) with GEN as reference category.
Therefore, we have three dummy explanatory variables in connection with caste.
OBC = 1, if the person belongs to the Other Backward Classes.
0, otherwise
SC = 1, if the person belongs to the Scheduled Caste
0, otherwise
ST = 1, if the person belongs to the Scheduled Tribe
0, otherwise
Variables in Models for Incidence of Domestic Violence against Women
In order to study the incidence of domestic violence against women in the district of
Bankura we have formulated two econometric models depending on the measures of
women’s empowerment. Except the empowerment variables, the variables in the models
are same. In the models for domestic violence against women, the incidence of Domestic
Violence is the dependent variable. It has been denoted by DVIO.
101
DVIO (domestic violence against women): We have measured this variable by
considering whether the woman experienced at least any two episodes of hit, kick, slap,
beat etc. by her husband or other family member in the last six months or not. The
incidence of domestic violence is a binary variable as specified below.
DVIO = 1, if the woman is victim of domestic violence during the last six months.
0, otherwise
Explanatory Variables
First, we list those explanatory variables which have already been specified in the
models for the decision regarding family planning. These are the DOWEH (Degree of
Women’s Empowerment at Household Level), CIWEH (Composite Index of Women’s
Empowerment at Household Level), SAGEG(Spousal Age Gap), HEDU (Educational
Qualification of the Husband), TYFAMI (Type of Family), Household Occupation,
CULTI (1= Cultivation), Household Occupation, NONFARM (1= Non-farm self-
employment), HLAND (Household’s Landholding), DRATIO (Dependency Ratio) and
APCHIN (Annual Per Capita Household Income). The explanatory variables indicating
the community traits are DOWEC (Degree of Women’s Empowerment at Community
Level), CIWEC (Composite Index of Women’s Empowerment at Community Level),
DSHGM (Duration of SHG Membership) and the dummies for Caste. In addition to
these variables, we have included some other variables in the individual/household
characteristics as specified below.
DURM (Duration of the Married Life): It is the chronological age of the woman after
her marriage and is measured in years.
HIMEDU (Highest Education among Male Household Members): It is the education
level of that male member of the family who attended the formal educational institution
for maximum period. The number of years that male member attended the educational
institution has been considered for measuring this variable.
DOW (Dowry Given at Marriage): It is captured the fact that whether the natal family of
the woman has given any kind of dowry (financial or physical assets) to her in-laws at or
before marriage It is a binary response variable as specified below.
102
DOW = 1, if the natal family of woman have given dowry.
0, otherwise
PMDOW (Post Marriage Dowry Demand): It is also a response variable stating whether
or not the in-laws of the woman has claimed for assets after marriage from her natal
house. Therefore,
PMDOW = 1, if the natal family of woman have given dowry after marriage.
0, otherwise
ADDIC (Addiction of the Husband): By addiction of the husband of the representative
woman we mean whether or not the husband is addicted to any kind of tobacco or
alcohol or both.
ADDIC = 1, if the husband is addicted to any kind of tobacco or alcohol or both.
0, otherwise
Variables in Models for Child Education Expenditure
Based on the different measures of empowerment we have formulated two separate
econometric models. These have been specified as Model-3A and Model-3B. In both the
models log of child education expenditure as proportion to household income in the last
year has been considered as the dependent variable. Except the empowerment variables,
the variables in Model-3A and in Model-3B are same.
LEDEX (Ln of Child Education Expenditure as Proportion to Annual Household
Income): It is the logarithmic value of the percentage of child education expenditure to
annual household income. We have already explained its measurement in section 3.3.3.
This is the dependent variable in the models for child education expenditure for the
women in Bankura district. No doubt it is a metric variable.
Explanatory Variables
First, we pass on the list of explanatory variables affecting child education expenditure,
which have already been specified in the previous models. These are the DOWEH
(Degree of Women’s Empowerment at Household Level), CIWEH (Composite Index of
103
Women’s Empowerment at Household Level), HEDU (Educational Qualification of the
Husband), HIMEDU (Highest Education among Male Household Members), TYFAMI
(Type of Family), Household Occupation, CULTI (1=Cultivation), Household
Occupation, NONFARM (1=Non-farm self-employment), HLAND (Household’s
Landholding), DRATIO (Dependency Ratio) and APCHIN (Annual Per Capita
Household Income). The explanatory variables indicating the community traits are
DOWEC (Degree of Women’s Empowerment at Community Level), CIWEC
(Composite Index of Women’s Empowerment at Community Level), DSHGM (Duration
of SHG Membership) and the Caste dummies. In addition to these variables we have
included one more variable in the individual/ household characteristics namely Highest
Education among Female Household Members (HIFEDU)
HIFEDU (Highest Education among Female Household Members): The highest number
of years in the educational institution, attended by any female member of the family, has
been considered as the highest education among female household members. So,
HIFEDU is a quantitative variable measured in year.
Variables in Models for Women’s Empowerment
In accordance with the two alternative measures of empowerment we have specified two
models for empowerment at the household level and two for empowerment at the
community level. We have defined four empowerment variables. For each of these
empowerment variables a regression model has been formulated. We have inserted the
same independent variables in each model for studying women’s empowerment at the
households and at the community level. We now specify them as follows.
Explanatory variables reflecting Individual/household level Characteristics
First, we mention the explanatory variables which have already been specified in the
previous models. These are EDU (education level of the woman), TYFAMI (Type of
Family), Household Occupation, CULTI (1=Cultivation), Household Occupation,
NONFARM (1=Non-farm self-employment), HLAND (Household’s Landholding),
DRATIO (Dependency Ratio), APCHIN (Annual Per Capita Household Income),
HIMEDU (Highest Education among Male Household Members) and HIFEDU (Highest
Education among Female Household Members). The explanatory variables indicating
the community traits are DSHGM (Duration of SHG Membership) and the Caste
104
dummies. In addition to these variables we have included some more variables as
follows.
Age (AGE): Age means simple the physical age of the person, counted by years.
However, in different stages of life, women play different roles and duties in their
households and in community. To gage the impact of age of women at different phases
of life on her empowerment at the household level and at the community level we have
divided them into four age groups: AGE1, AGE2, AGE3 and AGE4. Age of women in
our study is a categorical variable with AGE4 as the reference category. Therefore,
AGE1 = 1, if age of the woman is below 25 years
0, otherwise
AGE2 = 1, if age of the woman is into the range 25-35 years
0, otherwise
AGE3 = 1, if age of the woman lies in the range 36-45 years
0, otherwise
Occupational Status of the Woman: Occupation of a woman means the work that the
individual do on the most of the time of a day in general. We divide the occupation of
the women into three categories such that Homemaker, (HM) Wage labour (LAB) and
self-employed or service holder (SELF) with home maker category as reference
category. The women, who basically are engaged in household job or in field agricultural
activity without payment, belong to the category of home maker. The women who work
for wages or crop share in the agricultural sector or do household job for wage belong to
wage labour group. Women, who are mainly involved in business, manufacturing
activity, provide services in organized sector, work for wages or commission in the non-
agricultural unorganized sector belong to self-employed category. Thus we have
LAB = 1, if the woman belongs to wage labour group
0, otherwise
SELF = 1, if the woman earns from self-employment or from service
0, otherwise
105
PINC (personal income of the woman): It is amount of the money, measured in rupee
which the woman earns from her personal activity per month in average.
AFCT (Access to Formal Credit): It is a qualitative dichotomous variable indicating
whether or not the woman has access to formal credit. Therefore,
AFCT = 1, if the woman has access to formal credit.
0, otherwise
3.7. Specification of Hypothesis
In order to study the empowerment and related issues of rural women in the district of
Bankura we have already specified some models that to be estimated empirically. In this
section, we propose the relevant hypotheses respective to those models.
3.7.1. Hypotheses relating to the Model for Decision regarding Family Planning
In the model for decision regarding family planning, our purpose is to explain how the
household and community level empowerment of women along with other socio-
economic factors affect the probability of taking family planning decision. The probit
models that we have specified will enable us to test how the several explanatory
variables affect the probability of taking family planning decision. The relevant models
in this particular context are Model-1A and Model-1B and relevant hypotheses related to
these models are set below.
Hypothesis-1
The empowerment of women has some positive effect on the decision regarding family
planning. That is, 0β1
and 0β1C (refer to equation 3.5.1 and 3.5.2). As the women
become more and more empowered they can realise the benefits of having small family.
Empowered women do not take their children as their livelihood security of old age.
Rather they become more conscious about their children health and education. So, we
can say that the level of household empowerment is positively related with the
probability of taking decision regarding family planning.
106
Hypothesis-2
Women having two children with at least one female child are less likely to adopt family
planning. That means, 00,0β 432 and (refer to equation 3.5.1 and 3.5.2). It
is expected that if the household prefers male child compared to female, then the couple
under the household will not adopt family planning.
Hypothesis-3
Decisions regarding family planning are positively related with the age of the women at
marriage. Age of the women at marriage positively affects the probability of taking
family planning decision. That is, 0β5 (refer to equation 3.5.1 and 3.5.2).
Hypothesis-4
It is convention in Indian culture that a husband should be older than his wife. The
greater the difference of age between husband and wife the lower will be the women’s
say about family planning decision. The probability of taking family planning decision
by women is expected to be lower for larger age gap of spouses. That is, 0β6 (refer to
equation 3.5.1 and 3.5.2).
Hypothesis-5
Education of the women has positive impact on the decision regarding family planning.
That is, 07 (refer to the Model-1A and Model-1B).
Hypothesis-6
Education of the husband has a positive impact on the decision towards family planning
by woman. The probability of taking family planning decision by woman is directly
related with the age of the person. That means, 08 (refer to equation 3.5.1 and 3.5.2).
Hypothesis-7
Type of family, which the woman belongs to, affects the decision towards family
planning by women. The probability of taking family planning decision by women is
expected to be higher for the women of the nuclear family compared to those of the joint
family. That means, we expect 0β9 (refer to equation 3.5.1 and 3.5.2).
107
Hypothesis-8
The main occupation of the household is likely to influence the probability of taking
family planning decision by woman. The probability of taking family planning decision
is higher for the women of cultivator family and non-farm family than other. If the
family shifts from the wage labour class to the cultivator class or to the non-farm self-
employed family, the probability of taking family planning decision will increase.
Specifically, we like to test 0β10 and
0β11 (refer to equation 3.5.1 and 3.5.2).
Hypothesis-9
We can expect that the larger the size of landholding the larger will be the probability of
taking family planning decision. That is, 0β12 (refer to equation 3.5.1 and 3.5.2).
Hypothesis-10
Dependency ratio is expected to be negatively related with the probability of taking
family planning decision by women. That is, 0β13 (refer to equation 3.5.1 and 3.5.2).
Hypothesis-11
Annual per capita family income has a positive impact on the family planning decision
by women. That means that we like to test 0β14 for Model-1A and for Model-1B.
Hypothesis-12
Like the household level empowerment of women the community level empowerment of
women has some positive effect on the decision towards family planning. That is,
0β15 and 0β15C (refer to equation 3.5.1 and 3.5.2). Higher community
empowerment means women have larger mobility in the society. They can easily
identify the benefits of small family and troubles of having large family. So the women
having higher community level empowerment want to have a small family.
Hypothesis-13
The duration of SHG-membership increases the probability of taking family planning
decision. That is, 0β16 in equation 3.5.1 and 3.5.2. Being the member of SHG, a
woman becomes more conscious and gets in touch with other people. They understand
108
the importance and advantages of family planning. So the probability of taking family
planning decision increases with the increase of the duration of SHG-membership.
Hypothesis-14
The probability of taking family planning decision by woman is expected to be lower for
the people of OBC, SC and ST compared to that for the people of general caste. That is,
0 and 0,0β 191817 ββ (refer to equation 3.5.1 and 3.5.2).
3.7.2. Hypotheses relating to the Model for Incidence of Domestic Violence
In the model for the incidence of domestic violence, our purpose is to explain how the
empowerments of women along with other explanatory variables affect the probability of
suffering from domestic violence. This section presents the hypotheses logically. The
relevant models in the particular context are Model-2A (equation-3.5.3) and Model-2B
(equation-3.5.4) and relevant hypotheses related to these models are stated below.
Hypothesis-1
The probability of sufferings of women from the incidence of domestic violence is
negatively related with the household level empowerment and with the community level
empowerment. It means that we like to test 01 and 015 (refer to equation 3.5.3)
and we test 01 C and 015 C (refer to equation 3.5.4).
Hypothesis-2
Duration of married life of the woman is negatively related with the incidence of
domestic violence against her. That is, 02 (refer equations 3.5.3 and 3.5.4).
Normally, with the increase in age women become more experienced and their
influences in the family increase. Therefore, the duration of married life of the woman
inversely affects the probability of sufferings from the incidence of domestic violence.
Hypothesis-3
Spousal age gap is directly related with the occurrence of domestic violence against
women. The higher the age gap the higher will be the occurrence of domestic violence
against women. That is, 03 (refer to equations 3.5.3 and 3.5.4).
109
Hypothesis-4
Husband education has a negative impact on the occurrence of domestic violence. With
the increase in the level of education of husband, the probability of occurrence of
domestic violence against his wife decreases. That is, 04 (refer to equation 3.5.3
and 3.5.4). This is self-explanatory.
Hypothesis-5
Higher education of the male family member is likely to reduce the occurrence of
domestic violence. That is, higher education among the male person in family reduces
the probability of occurrence of domestic violence against women. That is, 05 (refer
to equation 3.5.3 and 3.5.4).
Hypothesis-6
The probability of occurring violence against women within family is expected to be
lower for the women of the nuclear family compared to those of the joint family. That
means, we expect 06 (refer to equation 3.5.3 and 3.5.4).
Hypothesis-7
It is very difficult to say whether the women of wage labour family suffer more from
domestic violence in contrast to the women of cultivator family or the non-farm self-
employed family. That is why, we want to test the hypothesis, 00 87 and (refer
to equations 3.5.3 and 3.5.4).
Hypothesis-8
There is no reason to expect any particular relation between the size of the household’s
landholding and the occurrence of violence against women. Therefore, we frame the
alternative hypothesis 09 (refer to equations 3.5.3 and 3.5.4).
Hypothesis-9
There is no confirmed relation between the dependency ratio and the probability of
occurrence of violence in the family against women. Therefore, we would like to test
010 (refer to the Models-2A and Model-2B).
110
Hypothesis-10
Higher the annual per capita family income lower will be the probability of the
occurrence of domestic violence against women. That means that we like to test
011 for equations 3.5.3 and 3.5.4. Higher per capita family income ensures higher
living and social standard of the family. So we expect that with the increase in the annual
per capita income, the probability of the occurrence of domestic violence against women
will decrease and vice-versa.
Hypothesis-11
We like to test whether the dowry system causes domestic violence in the area under
study. With this end in view, we make hypothesis that the women who have given any
kind of dowry suffer more from domestic violence from her in-laws. Not only that, we
expect the sufferings of women to increase if the post marriage dowry demand arises.
That is, we have set the hypotheses 012 and 013 (refer to the equations 3.5.3
and 3.5.4).
Hypothesis-12
The frequency of the sufferings of women from domestic violence and drug addition of
their husbands are directly related. That is why, we are interested to test 014 (refer to
the equations 3.5.3 and 3.5.4).
Hypothesis-13
There is a negative relation between the occurrence of domestic violence against women
and the duration of SHG-membership. That is, 016 (refer to the equation 3.5.3 and
3.5.4). In fact being the member of self-help group a woman becomes more conscious
about their rights and about their own wellbeing along with the wellbeing of the family
and the society.
Hypothesis-14
The women who come from the OBC, SC and ST family are more likely to suffer from
domestic violence compared to the women of the general caste. That is,
0 and 0,0 191817 (refer to the equation 3.5.3 and 3.5.4).
111
3.7.3. Hypotheses in Connection with the Child Education Expenditure
In this sub-section we would formulate a set of hypotheses stating the relation between
children’s education expenditure and women’s empowerment along with several other
determinants.
Hypothesis-1
Household as well as community level empowerment of a woman has a positive impact
on her children’s education expenditure. Therefore, the higher the empowerment of
women the higher will be the expenditure as proportion to household income for their
children’s education. So, we expect 01 and 011 in equation 3.5.5 and 01 C
and 011 C in equation 3.5.6.
Hypothesis-2
The higher the education level of the father, the higher will be the share of educational
expenditure in household income for children in the family. That is, 02 α (refer to
Models-3A and Model-3B).
Hypothesis-3
The educational expenditure out of family income is directly related with the highest
male education and highest female education in the family. That is, 0 and 0 43
(refer to Model-3A and Model-3B). Like many other decisions, the decision of how
much to spend in children education depends on the education level of the other family
members. An educated person better understand how much and for which stream of
education they would spend. It is natural to expect that the highest male education and
highest female education have favourable effect on the educational expenditure of the
family.
Hypothesis-4
Women who belong to the nuclear family spend more on the education of their children
than the women who belong to the joint or extended family. Thus we say that if a
woman moves from the joint family to nuclear family, she will spend more. That is,
05 (refer to equations 3.5.5 and 3.5.6).
112
Hypothesis-5
A mother of cultivator family or non-farm self employed family spends more on their
children education compared to a mother of wage labour family. That is,
00 76 and (refer to equations 3.5.5 and 3.5.6). In wage labour family more heads
means more income. Moreover, most of these families suffer from poverty. So they
concentrate more on their bread rather the education of their children. That is why, in
wage labour family children are also occupied as wage labour.
Hypothesis-6
Household’s landholding has positive effect on the children’s education expenditure. The
income share in children’s education expenditure increases as the size of the household
agricultural land holding increases. That is, 08 (refer to equations 3.5.5 and 3.5.6).
This is self-explanatory.
Hypothesis-7
Dependency ratio in the family affects negatively the educational expenditure.
Therefore, the expenditure for education of children and dependency ratio in the family
varies inversely. That is, 09 (refer to equations 3.5.5 and 3.5.6).
Hypothesis-8
As annual per capita family income increases share of child educational expenditure in
income increases. That means that we like to test 010 (refer to equations 3.5.5 and
3.5.6).
Hypothesis-9
The duration of membership of a woman in a SHG are expected to affect the expenditure
for their children education positively. As the number of years engaged in SHG
increases, the education expenditure for their children also increases. It implies
that 012 in equation 3.5.5 and 3.5.6.
Hypothesis-10
Women belonging to the general caste are more likely to spend on children’s education
compared to the women belonging to the Other Backward Classes, Scheduled Caste and
113
Scheduled Tribe. Therefore, the spending for children education as proportion to family
income will be higher in the general caste families contrasted with the non-general caste
families. That is, 000 151413 and , (refer to equations 3.5.5 and 3.5.6).
3.7.4. Hypotheses relating to the Models of Women’s Empowerment
In order to assess the impact of several socio-economic-demographic factors on
women’s empowerment at the household level and at the community level we have
specified four regression models in accordance with the measure of empowerment. In
this sub-section we propose the hypotheses relating to the empirical models.
Hypothesis 1:
It is very difficult to assign any particular relation between the age of the woman and the
empowerment level of women. We expect that the women’s empowerment at the
household level and at the community level of age group below 25 years is lower than
that of the reference group. Again we can assume that age groups 25-35 years and age
group 35-45 years enjoy more empower compared to reference group (above 45 years).
It means that 0,0,0 321 and (refer to equations 3.5.7 and 3.5.8) and
0,0,0 321 and (refer to equations 3.5.9 and 3.5.10).
Hypothesis 2:
Education of women directly affects women’s empowerment at the household level and
at the community level. Here our hypothesis is 04 referring to equation 3.5.7, 3.5.8
and 04 in equations 3.5.9 and 3.5.10. Education makes a woman more conscious in
all phases and all aspects of life. It increases her political and legal understanding. It
improves the decision making power of woman at home and outside home. From all
these observations we expect that education enhance both the household and societal
empowerment of woman.
Hypothesis 3:
The occupation of a woman is likely to influence her empowerment. Therefore, the
empowerment is higher for wage labour and self employed women compared to home
114
makers. Specifically, we like to test 05 and 06 (refer to equation 3.5.7 and
3.5.8). We also like to test 00 65 and in equation 3.5.9 and 3.5.10.
Hypothesis-4
Average Monthly Personal Income of the woman affects her empowerment level
positively. That is, 07 refer to equation 3.5.7, 3.5.8 and 07 in 3.5.9 and 3.5.10.
Hypothesis-5
There is a positive relation between the access to formal credit and the empowerment of
women. That is, we test 08 referred to equation 3.5.7 and 3.5.8 and 08 in and
3.5.9 and 3.5.10.
Hypothesis-6
Women belonging to nuclear family enjoy more empowerment compared to the women
belonging to joint family. This means 09 and 09 (refer to equations 3.5.7, 3.5.8,
3.5.9 and 3.5.10).
Hypothesis-7
Dependency ratio in the family is negatively related with the empowerment of woman.
We like to test 010 (refer to equations 3.5.7 and 3.5.8). We set alternative
hypothesis 010 in equations 3.5.9 and 3.5.10.
Hypothesis-8
We cannot confine to any definite relation between the annual per capita family income
and the empowerment of a woman of that family. In this case we test the hypothesis
011 (refer to equations 3.5.7 and 3.5.8) and 011 (refer to equations 3.5.9 and
3.5.10).
Hypothesis 9
Women who belong to the cultivator family or non-farm self-employed family has more
empowerment in compare to the women of wage labour family. That is,
115
00 1312 and in equations 3.5.7, 3.5.8 and 00 1312 and in equations 3.5.9
and 3.5.10.
Hypothesis-10
It is very difficult to identify the direction in which the empowerment of women move
when the size of the household’s landholding increase or decrease. In this case we test
the hypothesis 014 and 014 (refer to equations 3.5.7, 3.5.8, 3.5.9 and 3.5.10).
Hypothesis-11
Highest male education and highest female education in the household other than the
respondent is likely to accelerate the empowerment level of the woman in the household
as well as in the society. Therefore, the household and societal empowerment of women
varies directly with the highest male education and highest female education in the
household. Specifically, we like to test 015 , 016 and 015 , 016 (refer to
equations 3.5.7, 3.5.8, 3.5.9 and 3.5.10).
Hypothesis-12
The duration of SHG membership has positive impact on the empowerment of woman.
The higher the number of years of involvement of a woman in a SHG the higher will be
her empowerment at the home and at the society. This means 017 and 017 (refer
to equations 3.5.7, 3.5.8, 3.5.9 and 3.5.10).
Hypothesis-13
The empowerment of general caste women is expected to be higher for the women of
OBC, SC and ST, i.e. 0 and 0,0 201918 (refer to equation 3.5.7 and 3.5.8). We
also examine 0 and 0,0 201918 in equations 3.5.9 and 3.5.10.
3.8. Methodology of Data Collection
An adequate and reliable source of data is the primary ingredient of an empirical study.
Without an adequate and reliable data set, any type of empirical analysis and its results
would be falsified and will convey a wrong message to the future researchers and policy
makers and thereby the purpose of the empirical study will be lost. That is why, before
116
going to the any type of empirical analysis we need to have a set of adequate and reliable
data, which are collected following a scientific methodology. Keeping this point in
mind, we have collected a set of primary data from the district of Bankura in West
Bengal during 2012-2013. In this section we like to explain the methodology of data
collection and present the nature and scope of the data and finally present the procedure
of the diagnostic check for the sample size that would be used to factor analysis and
estimate the econometric models as specified in the above sections.
3.8.1. Sampling Design
In order to carry out the study of women’s empowerment we have considered the case
study of the district of Bankura in West Bengal. This district belongs to the Jangalmahal
(West Midnapore, Bankura and Purulia) in West Bengal. It has some distinct history and
cultural norms as mentioned in the chapter one. We have selected Bankura district
purposively as our study area. It is already known that the district has three sub-divisions
and twenty-two blocks. For conducting the sample survey we have followed four stages
stratified mixed sampling procedure. First of all, two blocks, namely Kotulpur and
Chhatna have been selected purposively. Of which Kotulpur block is relatively
developed and Chhatna block is relatively underdeveloped area in the district of
Bankura. This constitutes the first stage of our sampling. In the second stage of the
sampling we have randomly selected two Gram panchayets from Chhatna block and
three from Kotulpur block. Later on, we have chosen two (four villages from Dhaban
Gram panchayet) villages from each of the sample Gram panchayet. In total twelve
villages have been selected for our empirical study. Finally, after making a pilot survey
for each village, sample households are selected randomly from the sample villages. This
completes the fourth stage of the sampling design. It should be noted that number of
households chosen from each village are not equal. It depends on the total number of
households and other socio economic characteristics of the villages. Therefore, we have
designed a multi-stage sampling procedure which is also a combination of both
purposive and random sampling to take the advantages of the both. It may be looked as a
multi-stage stratified random sampling. Primarily, we have surveyed more or less six
hundred households and interviewed at least 611 persons; of them we have recorded the
relevant information of 580 households/persons in our data sheet. Data of some
households are rejected due to incomplete, insufficient, or absurd information and for
117
maintaining standard sample size suitable for factor analysis and econometric
estimations.
3.8.2. Profile of the Sample Areas
It has been already mentioned that our sample is constituted by 580 households residing
at villages belonging to two blocks in the district of Bankura. In this sub-section we
present an overview of the characteristics of the sample villages. Almost all the sample
villages are remote in terms of the access to well transport facility, health facility,
banking facility, job opportunity etc. In our sample we have selected three villages
namely Gopalpur, Sidabari and Meghkata, where scheduled tribe community forms the
major segment of population. In table-3.8.1 numbers and distribution of sample
households corresponding to each village have been presented. A large number of
Muslim households reside at Sarisadighi and Hati villages. Majority of the households at
Ghatdighi, Dhaban and Tegharia belongs to scheduled castes community. There is no
one village with any type of Bank branch.
Table 3.8.1. Area Specific Distribution of the Sample Households
Blocks GramPanchayets Villages Total
Households
Scheduled
Caste
Households
Scheduled
Tribe
Households
Sample
Households
Kotulpur
Madan
Mohan Pur
Hati 200 100 0 50
Sundarchack 200 80 0 49
Sihar Masinapur 260 80 20 60
Sihar 750 150 90 50
Kotulpur Ghatdighi 750 350 0 60
Sarishadighi 110 60 0 40
Chhatna
Dhaban
Dhaban 550 265 0 70
Tegharia 250 50 30 25
Sidabari 175 25 30 27
Gopalpur 265 55 130 44
Jamtora
Meghkata 120 0 90 60
Chitora 195 40 40 45
Total 580
Source: Author’s own field survey area during 2012-2013
118
There is a primary agricultural credit Samabay Samity at few sample villages. We have
taken different size of villages in terms of the number of total households residing at the
villages. In this sense, Sihar is the largest village and Sarisadighi is the smallest one
among the sample villages. But sample households selected from Sihar is not highest
among the number of households selected from other sample villages because a large
section of the households in this village lies in a homogeneous group. In almost all the
villages, the SHG based micro financing has been functioning but their tenure and
outreach of activities differ from village to village. Most of the groups have been formed
by women. This program includes the village women in the main stream development
process making their financial inclusion and inculcates empowerment. In few villages
women have participated in the NREGA program for supporting their families.
3.8.3. Nature and Scope of Data
We had prepared a structured questionnaire based on the objectives and necessity of our
empirical study. For each village first we have conducted a pilot survey regarding the
village level information. Then, the required data have been collected directly from the
representative women from households. The purposive information has been collected
through personal interview method. A woman from each sample household has been
interviewed personally in a face-to-face contact to collect relevant information. The
conversation was conducted in the local language. Therefore, we can claim that our data
set is purely primary in nature.
In order to study the nature causes and consequences of women empowerment we have
conducted exhaustive household survey and gathered the information regarding the
different indicators of empowerment at the household level and at the community level
as mentioned in section 3.2. Information regarding different household welfare has been
collected to assess to impact of empowerment. Particularly data for several issues on
domestic violence and on decision regarding family planning have also been collected.
We have covered household income and expenditure pattern on food nutrition, on fuel
and energy, on health care and on child education. We have also collected information
regarding some selected socio economic and demographic characteristics of the
respondent women, households and village like social status, age, educational
background of the woman and other household members, occupation of the woman and
occupational structure of the family, landholding, annual family income, dependency
119
ratio, SHG-membership status, caste, religion, distance of household from bank, school,
hospital etc. We have divided categorically all the relevant information into individual,
household and community characteristics. The respondents are not always smart enough
to inform actual figure of the required data. So, we have logically tried to make a
generalization. It is natural that in any type of data collection there may be some sort of
discrepancy or personal bias. In our case we have tried to minimize this problem through
cross checking.
3.8.4. Diagnostic Check for the Sample Size
In order to study women’s empowerment at the household level and at the community
level we have planned to apply factor analysis. So, we should check the factorability of
our data set. The primary condition for applying the factor analysis is that the variables
under consideration, indictors of women’s empowerment in our case, are significantly
correlated with each other. There are two popular tests for diagnostic check of the data
set to be used in factor analysis. One is Bartlett’s test of sphericity and other is the
Kaiser-Meyer-Olkin measure of sampling adequacy.
In order to test the null hypothesis that the correlation matrix of the variable under
consideration is an identity matrix we have used Bartlett's Test of Sphericity based on
the test statistic
2
)1(2
1
2 ~])/1ln()][ln/212(6
1)1[(
kkjIkkDkkn where, k stands for
number of variables, jI denotes jth eigenvalue of D.
Kaiser-Meyer-Olkin measure of sampling adequacy has been used to compare the
magnitudes of the observed correlation coefficients in relation to the magnitudes of the
partial correlation coefficients of the variables with the help of the formula
22
2
ijij
ij
r
rKMO
where, kijr ,....2,1.ij
The range 0.8 to 1.0 of KMO statistic indicates a commendable degree of common
variance. If KMO statistic lies in the range 0.6 to 0.8 we can say that the degree of
120
common variance is mediocre. The value of KMO statistic below 0.6 reveals a miserable
degree of common variance and the researchers should not use the factor analysis. This
methodology suggests that before going to Principal Component Analysis for
empowerment indicators we have to conduct the KMO test of the data set of the
indicators of women’s empowerment. If we find KMO value greater or equal to 0.6 we
apply PCA and otherwise not. We should, therefore, obey this rule in our study of
women’s empowerment.
Once our data set of the indicators of women’s empowerment passes this diagnostic test
we can apply Principal Component Analysis for extracting the factors of women’s
empowerment and their respective weight. It will help us measure a composite index for
women’s empowerment at the household level and at the community level.
3.9. Conclusion
In this chapter we have explained the women’s empowerment conceptually and
quantitatively. We have proposed two alternative methodologies for quantifying the
women’s empowerment at household level and at the community level for the sample
women. The frameworks for studying the impact of women’s empowerment, on
household welfare denoted by the attitudes regarding family planning, domestic violence
against women and children education expenditure have been presented. We have
presented the regression specification of our analytical framework of the issues of
women’s empowerment in the district of Bankura followed by the underlying hypotheses
in the respective models. Given the model, methodology and data, we present and
discuss the empirical findings of the issues relating to women’s empowerment in the
district of Bankura in chapter four and chapter five.
121
Chapter Four __________________________________________________
Component Analysis of women’s empowerment
4.1. Introduction
In chapter one we have chalked out some specific research issues related to women’s
empowerment in the district of Bankura that are to be addressed in this dissertation.
Chapter two has exposed the motivation and backdrop of our empirical investigation.
Against this backdrop, in chapter three we have formulated the relevant working models
and set our hypotheses related to the study of women’s empowerment. In this chapter we
are going to discuss the descriptive statistics of our sample women. We analyse the
components of empowerment of the sample women and inter correlations among several
variables in this chapter. For this purpose we have conducted a self-designed household
survey during the period 2012-13.
The route of journey of this chapter has been designed as follows. In section 4.2 we have
presented the descriptive statistics of surveyed households along with its three sub-
sections. In sub-section 4.2.1 we have explained the categorical characteristics of the
sample households followed by the discussion on the categorical characteristics of the
respondents. We have presented the socio-economic variables and demographic
quantitative features of the surveyed population in sub-section 4.2.2. Section 4.3 deals
with the discussion on the relevant indicators of women’s empowerment at the
household level and at the community level. Section 4.4 elucidates findings related to
women’s empowerment indices. It has three sub-sections. Sub-section 4.4.1 deals with
the outcomes of Principal Component Analysis of the household level empowerment of
the women. In sub-section 4.4.2 we have presented the results of the Principal
Component Analysis of the community level empowerment of the women. We have
122
presented the descriptive statistics of the empowerment indices in section 4.4.3. Section
4.5 has explained the profile of inter-correlation among different socio-economic-
demographic and empowerment variables of the sample women/households. We have
concluded this chapter in section 4.6.
4.2. Descriptive Statistics of the Surveyed Households
Nobody would deny that in an empirical study it is very much important to look into the
nature of the socio-economic and demographic characteristics of the surveyed
households and respondents. In order to carry out this empirical analysis we have
conducted a primary level household survey in selected regions of Bankura district, West
Bengal. From this survey and field observations we have collected information about
socio-economic and demographic characteristics of households of our study region. This
empirical study is based on the information of 580 women/households in the district of
Bankura surveyed during the period 2012-13. During the course of our household survey
we have observed that the households surveyed for our study are of different categories
with respect to their family composition, their main source of livelihood, with respect to
their caste and religion. There is also some heterogeneity among the sample
women/households in terms of household income, landholding, family size, number of
child, dependency ratio, household expenditure pattern, social caste and community etc.
With this end in view, we have classified the socio-economic and demographic
characteristics of the surveyed households and women into two types of variables – one
is the categorical variable and other is the quantitative variable. In this section with its
two sub-sections we are going to present the descriptive statistics of the surveyed women
and their households.
4.2.1Categorical Characteristics of the Sample Households/Individuals
In this sub-section we present the categorical characteristics of the sample households
and that of the sample women. Table-4.2.1 depicts the frequency and percentage
distribution of the categorical variables related to households and table-4.2.2 depicts the
percentage distribution of the categorical variables related to the sample women.
The table-4.2.1 shows that out of 580 sample women 475 (81.90 %) women belong to
nuclear family, whereas, only 105 (18.10%) women belong to extended family. It
indicates that most of the sample women in rural Bankura live with their own and next
123
generations. In term of household occupation we find that 44.31% of sample households
are engaged in cultivation in their own land. Our study covers two community blocks,
namely, Kotulpur which has cannel and tube well based irrigation facility and Chhatna
which has no developed irrigation facility for cultivation. As a result we find that farmers
of Kotulpur block can engage themselves in agricultural works throughout the year while
most of the farmers of Chhatna block stay unemployed seasonally. The main occupation
of one third of the surveyed households is wage labour. A major section of this wage
labour is engaged in agriculture. Like the farmers, agricultural labourers in Chhatna
block remain unemployed seasonally and agricultural labourers in Kotulpur get job
almost whole year. We have observed that the wage rate of agricultural labourer in
Chhatna is lower than that of the agricultural labourers in Kotulpur block. A large
number of respondents have reported that the program of NREGA recently reduces the
length of seasonal unemployment and makes a wage hike for the agricultural labourers in
the area under study. However, farmer households have reported that NREGA program
sometimes creates the shortage of labour at the sowing and harvesting time and as a
result it makes the cultivation costly. Moreover, a good number of the wage labourers in
our surveyed area works at brick field, sand lifting field, stone crashing and
constructional activity. Of the sample households 14.31% earns their livelihood mainly
from self-employment. Our field observations show that most of the self-employment
activities of our sample households are either grocery shop or some traditional ancestry
occupations like, worshipping, barbering, pottery, cobbler, carpentry, etc. Only seven
percent of the sample households belong to service sector by occupation. It is pointed out
that majority of the service holders in the area under study are engaged in education
sector or in defense sectors.
Our study covers four categories of castes. We find from the table that 31.21% of sample
households belong to general caste, 23.28% to other backward classes, 33.62% to
scheduled caste and 11.90% to scheduled tribes. As majority of the households in the
district of Bankura are under Hindu religion the sample survey has covered most of
Hindu households. This table shows that 97.07% of households belong to Hindu religion
and the remaining households belong to Muslim (2.93%). In accordance with the
notified poverty line (BPL card holders) our study observed that nearly 41.90% of
sample households live above poverty line whereas 58.10% below the poverty line.
124
Table-4.2.1 Distribution of the Categorical Variables of the Sample Households
Variable Category Frequency Percent Cumulative
Percent
Family Composition Extended 105 18.10 18.10
Nuclear 475 81.90 100.00
Household Occupation
Cultivation 257 44.31 44.31
Self-employment 83 14.31 58.62
Wage labour 198 34.14 92.76
Service 42 7.24 100.00
Social Caste of the
Sample Households
General Caste 181 31.21 31.21
Other Backward Classes 135 23.28 54.48
Scheduled Caste 195 33.62 88.10
Scheduled Tribe 69 11.90 100.00
Religion Muslim 17 2.93 2.93
Hinduism 563 97.07 100.00
Economic Status Above Poverty Line 243 41.90 41.90
Below Poverty Line 337 58.10 100.00
Accessibility to Adequate
Sanitation
No 398 68.62 68.62
Yes 182 31.38 100.00
Access to Safe Drinking Water No 405 69.83 69.83
Yes 175 30.17 100.00
Access to Advanced Cooking
Fuel
No 435 75.00 75.00
Yes 145 25.00 100.00
Whether any household
member is addicted in drug
(like smoking, tobacco, liquor,
etc.)
No 280 48.28 48.28
Yes 300 51.72 100.00
Access to Affordable housing No 203 35.00 35.00
Yes 377 65.00 100.00
Prevalence of Child death
before completing five Years
of Age
No 514 88.62 88.62
Yes 66 11.38 100.00
Source: Author’s computation based on primary data, 2012-13
Although several governmental and non-governmental sanitation programs have been
functioning for last one and half decades in the area under study, we observe that only
31.38 % of our sample households have accessibility to adequate sanitation. We find that
only 30.17% of sample households have accessibility to safe drinking water. A large
number of households have reported that they collect drinking water from their own or
neighboring house’s well or tube well where water is contaminated with huge iron and
other unhygienic minerals. Most of the women have not access to advanced cooking
fuel. Only 25% have accessibility to it. Our study shows that at least one member of
51.72% sample households is addicted to smoking, tobacco, liquor etc. However, we
find that most of the sample women are not addicted to any kind of drug. Only a few are
addicted to tobacco (gurakoo) and a handful of women smoke. In terms of affordable
125
housing defined as the per capita number of rooms in the house being greater or equal to
one, 377 households have accessibility to affordable house. They are 65% of our total
sample households and the remaining 35% have not accessibility to it. We find that there
is prevalence of child death before completing five years of age in 11.38% of sample
households in the area under study. It is greater than the rate of child death prevalence in
our country as a whole (Economic Survey, 2012-13).
Table 4.2.2 gives the percentage distribution of the individual categorical variables of
sample women. All sample women are of age eighteen or above eighteen years. We find
that out of 580 sample women 64 (11.03%) women are of age below twenty five years,
207 (35.69%) women belong to age group 25-35 years, 199 (34.32%) women belong to
age group 36-45 years and 110 (18.96%) women are of age above 45years. The age
distribution of the sample women shows that our study has covered all age groups of
women. It will help us study the nature and consequences of empowerment in different
age groups of women in Bankura district, W.B. In our sample most of the women are
married (92.93%). A very few proportion of the sample women are separated with
husband (0.52%). Our sample consists of 6.55% widow.
We find that nearly half of the sample women (49.7%) are exclusively homemaker. Our
field survey finds that 15.3% of sample women work in their family land together with
other family members. We have got 7.10% of our sample respondents as self employed.
One fourth of the sample women works as wage labours. Among the wage labourers
most of the women work in agricultural sector. In this sample only 3.6% of the women
are engaged in service sector. Note that among the service holders majority are serving
the education sector. One or two are engaged in health sectors. No one woman in our
sample is serving as administrator.
This table shows that 61.38% of the sample women have savings account in a bank,
whereas only 38.62% do not have bank account. It means that a large part of the rural
women in the district of Bankura are financially excluded. We find that 43.97% are
member of self-help group and 56.03% are not member of any self-help group.
Therefore, the SHG-centric microfinance program has covered a broads section of the
rural women in the inclusive development process. In our sample 40.69% of women
have their own mobile phone.
126
Table-4.2.2 Percentage Distribution of the Individual Categorical Variables
Variable Category Frequency Percent Cumulative
Percent
Age
<25 Year 64 11.03 11.03
25-35Year 207 35.69 46.72
36-45Year 199 34.32 81.04
45< Year 110 18.96 100.00
Marital Status
Divorce 3 0.52 0.52
Married 539 92.93 93.45
Widow 38 6.55 100.00
Personal Occupation
Homemaker 288 49.7 49.7
Farming 89 15.3 65.0
Self-employed 41 7.10 72.1
Wage labour 141 24.30 96.4
Service 21 3.6 100.00
Savings Bank Account Holder No 224 38.62 38.62
Yes 356 61.38 100.00
Self-Help Group Membership No 325 56.03 56.03
Yes 255 43.97 100.00
Mobile Phone Holder No 344 59.31 59.31
Yes 236 40.69 100.00
Dowry Given at Marriage No 179 30.86 30.86
Yes 401 69.14 100.00
Post Marriage Dowry Demand No 489 84.31 84.31
Yes 91 15.69 100.00
Experience of any kind of
Domestic Violence
No 194 33.45 33.45
Yes 386 66.55 100.00
Experience of Physical Violence
from Family members
No 253 43.63 43.28
Yes 327 56.37 100.00
Response toward Family Planning
Negative 236 40.69 40.69
Positive 344 59.31 100.00
Source: Author’s computation based on primary data, 2012-13
Most of the sample women (69.13%) have reported that their guardians gave dowry to
their bridegroom at the time of their marriage. It has been found that 15.69% of sample
women fulfil the post marriage dowry demand of their in-laws. Our study shows that
66.55% have the experience of any kind of domestic violence and 56.37% have the
experience of physical violence. This dissertation has tried to find out the nature and
causes of domestic violence of the rural women in the district of Bankura. We find that
59.31% have taken family planning decision. So, a large section of rural women in
Bankura district is still not conscious regarding the family planning. Therefore, the study
127
of the nature and causes of the response towards family planning in the area under study
is also a relevant issue and it has been addressed in this dissertation.
4.2.2 Socio-Economic and Demographic Characteristics of the Surveyed Population
This section describes the summary statistics of the quantitative variables in question.
Table-4.2.3 presents the demographic profile of sample women. Table-4.2.4A and table-
4.2.4B present the descriptive statistics of the socio-economic variables of our surveyed
households.
Table-4.2.3 Relevant Demographic Profile of the Sample Households
Family
Size
(Number)
Number
of
Children
Age
(Year)
Age of
Husband
(Year)
Spousal
Age
Gap
(year)
Duration
of
Marriage
(Year)
Age at
Marriage
Time
(Year)
Dependency
Ratio (%)
Mean 3.86 1.53 35.62 41.36 5.74 16.87 18.75 49.68
Median 4.00 2.00 35.00 40.00 5.00 17.00 19.00 50
Max. 8.00 5.00 65.00 72.00 22.00 55.00 42.00 100
Mini. 1.00 0.00 18.00 21.00 0.00 0.00 13.00 0
S D. 1.17 0.96 9.61 10.25 2.83 9.41 4.92 21.98
Skew. 0.29 0.23 0.50 0.47 1.52 0.62 0.25 -0.66
Kurtosis 3.30 3.02 2.99 2.95 6.79 3.36 6.94 2.95
Source: Author’s computation based on primary data, 2012-13
We find from table-4.2.3 that on an average each family has 3.86 members while
maximum number in a family is 8. This table shows that the half of the sample families
has more than 2 children. The average age of the respondents is 35.62 years. The Median
value of the age of sample women in the area under study is 35 years. The physical age
of the sample woman ranges from 18 years to 65 years. We find from this table that
average age of the husbands of the respondents is 41.36 years. The ages of the husbands
range from 21 years to 72 years. The average of the difference of ages between husband
and wife is 5.74 years, while the maximum age difference is 22 years. On an average
duration of married life is 16.87 years. The median value of the duration of married life
is 17 years. Our field survey has reported that the average age at the time of marriage is
18.74 years with a minimum age at marriage 13 years. Our household survey shows that
the average dependency ratio is 49.68%. This means that half of the family members in
average depend on other for their bread and butter.
128
Table-4.2.4A describes the socio-economic variables of the sample households. The
average annual per capita income of a family is Rs 13790.00. The median of annual per
capita income is Rs 10510.00. This means that the majority of the sample family lies
below poverty line. The maximum value of per capita household income of the surveyed
households is Rs 150000.00 and the minimum of that is Rs 3900.00. This implies that
sample families are basically poor. Moreover, some families earn nothing to live their
livelihood. This table shows that average monthly income of the surveyed women is Rs
594.00 with its maximum value Rs 35000.00 and minimum value Rs 00.00. The median
value personal monthly income is Rs 350.00. This implies that half of the sample women
earn less that Rs 350.00 in a month. It focuses that majority of the working women in the
district of Bankura is ill paid.
Table-4.2.4A Socio-economic Characteristics of the Sample Households
Annual
Per capita
Income
(Rs. ‘000)
Monthly
Personal
Income
(Rs. ’00)
Monthly
Expenditure
for
Education
(Rs. ’00)
Monthly
Expenditure
for Energy
& Fuel
(Rs. ’00)
Monthly
Expenditure
for Food &
Nutrition
(Rs. ’00)
Monthly
Expenditure
for Health
& Hygiene
(Rs. ’00)
Land
Holding
(Bigha)
Mean 13.79 5.94 2.79 2.77 20.11 2.50 2.64
Median 10.51 3.50 1.50 2.00 18.50 2.00 2
Maximum 150.00 350.00 30.00 20.00 54.00 50.00 16
Minimum 3.90 0.00 0.00 0.00 2.14 0.20 0
Std. Dev. 13.87 20.50 4.01 2.78 8.54 2.76 2.99
Skewness 5.44 12.09 3.04 2.47 1.16 9.68 1.53
Kurtosis 41.42 177.56 15.24 10.95 4.66 154.63 4.99
Source: Author’s computation based on primary data, 2012-13
Table-4.2.4A shows that on an average each family spends Rs 279.00 per month for their
children’s education. The average monthly expenditures for ‘food and nutrition’ and
‘health and hygiene’ are Rs 2011.00 and Rs 250.00. The average of household’s
agricultural land holding for all sample households (including landless and landed) is
2.64 bigha (1 bigha= 0.4 acre). Agricultural land holding is one of the major sources of
income for poor rural household. The household’s agricultural landholding ranges from
zero to sixteen bigha. Apparently, holding sixteen bigha of agricultural land is good in
size. But it is not the fact. It is good only if it is in the fertile part of our study area,
Kotulpur. It does not necessarily imply the well economic condition of the households if
it is on rocked and drought prone area like the block Chhatna. So, we should not
129
concentrate on its range. The median value of household’s land holding is 2 bigha. This
means that half of the surveyed households own less that 2 bigha of land. In terms of
landholding this study confirms that households belong to small and marginal farmers.
We can conclude from the income as well as expenditure pattern of surveyed households
that we are going to examine the empowerment status of women of poor society in the
district of Bankura.
Table-4.2.4B Socio-economic Characteristics of the Sample Households
Statistics
Education
of the
Respondent
woman
(Year)
Spouse
Educational
Gap (Year)
Husband’s
Education
(Year)
Highest
Education level
among
household
males(Year)
Highest
Education
level
among
females
(Year)
Duration of
SHG
membership
(month)
Mean 3.59 1.27 4.87 6.99 5.54 27.24
Median 1.00 0.00 5.00 8.00 6.00 0
Maximum 17.00 15.00 17.00 17.00 17.00 145
Minimum 0.00 -10.00 0.00 0.00 0.00 0
Std. Dev. 4.16 2.98 4.77 4.57 4.55 36.79
Skewness 0.74 0.52 0.40 -0.12 0.21 1.15
Kurtosis 2.43 5.22 1.98 2.23 2.07 3.46 Source: Author’s computation based on primary data, 2012-13
We now illustrate the table-4.2.4B. It describes some socio-economic variables of
sample households. We see from this table that the level of education of the respondent
women varies from zero to seventeen years but mean of the level of education is only
3.59 years. This finding indicates that the average of sample women cannot complete the
primary level of education (fourth class standard). Besides the education level of the
respondent women, we have recorded information about the educational levels of the
other members of the family with the intention of exploring the relation between
family’s educational background and women’s empowerment. This table shows that the
average level of education of husbands of the respondent women is 4.87 years.
Education level of the husbands ranges from zero to seventeen years in formal
educational institution. We have calculated that the average of the gap of education
between husband and wife is 1.27 years. The positive value of this gap implies that in
most of the cases husbands are more educated than their wives. The median value of
these differences is zero. This means that in fifty percent cases wives are equally
educated as their husband.
130
Table-4.2.2B also shows that mean of the highest educational qualification among the
male members of family in the surveyed families is 6.99 years. The mean of highest
female education other than the respondent woman is 5.54 years. The table exhibits that
average of the highest education of female is lower than that of male among the sample
families. It has been noted that average of the highest education level of male and female
are relatively higher than the average education level of the persons interviewed for the
study. Therefore, in terms of education, our sample households as well as respondents
are not forward enough. Our study has captured another socio-economic attribute of the
sample women i.e., whether she is a member of SHG-based microfinance program or
not. We have found that the majority of the surveyed women are member of SHG with
average duration of membership of two years and three months. The highest length of
membership of our sample women is twelve years. During the course of our field survey
we have observed that regarding the financial service the primary cooperative societies
have come forward compared to the government program (SGSY) and banking sector in
the area under study.
4.3. Frequency Distribution of the Indicators of Women’s Empowerment at
Household Level and at Community Level
In this study we have measured empowerment of women at household and community
level. To this end, we have set some indicators. In this section we will show the frequency
distribution of these indicators. Table-4.3.1A and table-4.3.1B show the percentage
distribution of the indicators of women’s empowerment at household level. We have
presented the percentage distribution of the indicators of women’s empowerment at
community level in table- 4.3.2A and in table-4.3.2B.
Table-4.3.1A illustrates the indicators of economic, political and socio-cultural
dimensions of women’s empowerment at household level. From the economic dimension
we see that that 71.9% respondents have control over their own income, 79.83% have
access to household resources. This table exhibits that 12.07% of the sample women bear
more than fifty percent of household expenditure, 46.38% bear less than fifty percent of
household expenditure and 41.55% bear no part of it. We see that 6.9% respondent
women take the saving and investment decision of their own; 79.66% take this decision
jointly with their husbands; 8.62% decide on this matter by consulting with other
member in the family and 4.83% cannot opine in this matter.
131
Table-4.3.1A Percentage Distribution of the Indicators of Women’s Empowerment
at Household Level Indicator Value Frequency Percent Cumulative (%)
Eco
no
mic
Dim
ensi
on
Whether she has control over her personal
income/asset (COPI) (Yes=1) 0 163 28.1 28.1
1 417 71.9 100
Whether she can have access to household
resources (ACHR) (Yes=1)
0 117 20.17 20.17
1 463 79.83 100
Proportion of household expenditure that she
bears (PHESB)(None = 0/ less than 50% = 1
/greater than 50%=2)
0 241 41.55 41.55
1 269 46.38 87.94
2 70 12.06 100
Who decide the use of saving/ loan? (DUSOL)
(Own =4 / with spouse = 3/ with other family
member =2 /other members =1)
1 28 4.83 4.83
2 50 8.62 13.45
3 462 79.66 93.1
4 40 6.9 100
Does she take part in the decision for selling or
buying asset for household? (SOBHA) (Yes=1) 0 53 9.14 9.14
1 527 90.86 100
Does she enjoy freedom in choosing her
occupation? (CHOU) (Yes=1) 0 377 65 65
1 203 35 100
Po
liti
cal
Dim
ensi
on
Do you get domestic support for your political
engagement? (DSPE)(Yes=1) 0 502 86.55 86.55
1 78 13.45 100
Does she know the name of local (panchayat
pradhan / councilor/ MP/ MLA)? (KLPL)
(Yes=1)
0 92 15.86 15.86
1 488 84.14 100
Whether she know the candidate of opposition
party in the last election. (KOPL) (Yes=1) 0 423 72.93 72.93
1 157 27.07 100
Did she cast her vote in the last election?
(CASVO) (Yes=1) 0 83 14.31 14.31
1 497 85.69 100
Does other influence her to cast her vote?
(INVO) (Yes=0) 0 46 7.93 7.93
1 534 92.07 100
So
cio
-Cu
ltu
ral
Dim
ensi
on
Does she regularly enjoy Radio, telephone, TV
and Newspaper? (ENTM) (Yes=1) 0 244 42.07 42.07
1 336 57.93 100
Whether she is free to move outside your home.
(FMOH) (Yes=1) 0 73 12.59 12.59
1 507 87.41 100
Do you want to educating your girl or other
girls in your household (GEDU) (Yes=1)
0 190 32.76 32.76
1 390 67.24 100
Whether or not she participates in local cultural
programs.(PLCP) (Yes=1) 0 298 51.38 51.38
1 282 48.62 100
Whether she want to send her child for earning.
(CLAB) (Yes=0) 0 348 60 60
1 232 40 100
Would you arrange the marriage of the girls
before their eighteen years old or support it?
(CMAR) (Yes=0)
0 402 69.31 69.31
1 178 30.69 100 Source: Author’s computation based on primary data, 2012-13
132
It is revealed that 90.86% of sample women take part in the decision of selling or buying
household asset and 35% of them enjoy the freedom in choosing their own occupation.
We have seen that in the case of some indicators women in the district of Bankura are
well-off but in many cases they are worse-off. Therefore, it is difficult to judge the
nature of economic empowerment of the sample women from the percentage distribution
of the indicators.
This study has considered five indicators to cover the political dimension of women’s
empowerment at the household level. Table-4.3.1A shows that only 13.45% of the
sample women get domestic support for their political engagement. We have come to
know from this table that 84.14% of the respondents know the names of local Panchayet
-pradhan/ MP/MLA where as only 27.07% know the name of the candidate of opposition
party. This table also discloses the fact that though 85.69% cast their vote in last election
and 7.93% has been influenced by other regarding the issue of casting of vote and whom
to cast vote. We see that there is a disparity regarding the favourable responses towards
political empowerment at the household level of our sample women.
Let us consider the socio-cultural dimension which has five criteria. This table shows
that 57.93% of respondent enjoy radio, television, telephone and newspaper regularly;
87.41% have the freedom to move outside of their home and 67.24% want to educate
their own girl child and other girl in their family. We find from this table that 48.62%
participate in local cultural programme. It has also been reported that 60% of the
respondents are ready to send their child for earning and 69.31% support the marriage of
girl before reaching the age of eighteen. These results portray a very awful picture of
poverty as well typical psychology about girl chid in our society.
We, now, turn to table-4.3.1B. It is an extension of the previous table. It shows the
frequency distribution of the indicators of familial/personal dimension and that of legal
dimension. The familial/personal dimension and legal dimension have five indicators
each. From our field survey, we have come to know that only 6.9% of the sample
respondents’ marriages were arranged on self decision. It has been reported that 79.83%
of respondents can articulate their personal problems to other family members. We have
observed that in 17.41% cases some other person/household members interfere during
the time of the interview. More than half of the respondents (53.45%) have told us that
133
they can independently decide on their child education, health, food etc. This table
shows that 63.97 % of sample women have decision making power regarding their
personal health, body and family planning.
Table-4.3.1B Percentage Distribution of the Indicators of Women’s Empowerment
at Household Level
Indicator Value Frequency Percent
Cumulative
Percent
Per
son
al /
Fam
ilia
l D
imen
sio
n
Whether her marriage is arranged or self selection.
(MATY) (Self selection=1)
0 540 93.1 93.1
1 40 6.9 100
Can she articulate her personal problem to other
family members? (ARTIP) (Yes=1)
0 117 20.17 20.17
1 463 79.83 100
Whether anybody interfere when she talk to
stranger. (INFE) (Yes=0)
0 101 17.41 17.41
1 479 82.59 100
Whether or not she can independently decide about
your child education, health, food etc.(INDEC)
(Yes=1)
0 270 46.55 46.55
1 310 53.45 100
Whether she has decision making power regarding
her personal health, body, and family planning
(PDEC) (Yes=1)
0 209 36.03 36.03
1 371 63.97 100
Leg
al D
imen
sio
n
Whether she know about the mechanism of justices
used in the locality (KAMJ) (Yes=1)
0 96 16.55 16.55
1 484 83.45 100
Does she think women and men get equal treatment
from this system? (TRMW) (Yes=1) 0 189 32.59 32.59
1 391 67.41 100
Whether she knows the laws and legislation
available in favour of women. (LAFW) (Yes=1) 0 197 33.97 33.97
1 383 66.03 100
Whether she know about the various kinds of
public services available in the locality. (PUBS)
(Yes=1)
0 238 41.03 41.03
1 342 58.97 100
Whether your marriage is registered or not(MARR)
(Yes=1) 0 558 96.21 96.21
1 22 3.79 100 Source: Author’s computation based on primary data, 2012-13
Let us now consider the indicators of the legal dimension. Table-4.3.1B shows that
83.45% of the sample women know about the mechanism of justice used in the locality.
This study finds that 67.41% of the respondents think that women and men get equal
treatment from this system of justice. It is revealed that 66.03% of respondents are
familiar with the laws and legislation available in favour of women and 58.97% know
the various kinds of public services available in their locality. This study unveils the fact
that the marriage of only 3.79% of the respondents is registered. The above information
regarding the indicators of empowerment tells us that in terms of some indicators women
134
in the district of Bankura are forward but in other cases they have huge lag. Therefore,
the percentage distribution of the indicators of women’s empowerment at the household
level is not sufficient to understand the overall nature of empowerment at the household
level. With this end in view, we have tried to make component analysis for analysing the
overall empowerment of the women at the household level.
Table-4.3.2A and table-4.3.2B show the percentage distribution of the indicators of
women’s empowerment at the community level. Table-4.3.2A shows the percentage
distribution of indicators of economic, political and socio-cultural dimension of women’s
empowerment at the community level. We find that 52.07% women earn money; 85.17%
have land or property of their own and 54.83% has access to formal saving, insurance,
loan etc. Half of the sample women (50.17%) have access to formal education and
training as per their own requirement. About 4.14% report that other person threats them
to evict from property. Table-4.3.2A shows that the occupation of 51.55% of the sample
women is secured. It roughly indicates that almost half of the sample women have no
livelihood security.
The indicators of political dimensions of women’s empowerment at community level
reveal that one fifth of the respondents are involved in active politics and 8.10%
introduce themselves as volunteer of any political party. It has been reported that 42.93%
willingly attend political gathering and 3.62% contest any kind of vote as a
representative at least once in life. We have found that one third of our sample member
act as leader of an organization.
Frequency distribution of the indicators of the socio-cultural dimension tells that 47.07%
of the respondents are member of any social organization. About 22.07% of sample
women think that they can influence the election or selection of leader of these
organization or group. This table shows that 63.79% participates in community activity.
We know from the table-4.3.2A that almost all of the sample women (95.34%) know the
location of nearest post office, school, hospital, club, vegetable market etc. This study
reveals that 15% of the sample women feel exclusion from participation in some
community activities because of being women.
135
Table-4.3.2A Percentage Distribution of the Indicators of Women’s Empowerment
at the Community Level
Indicator Value Frequency Percent
Cumulative
Percent
Eco
no
mic
D
imen
sio
n
Whether she is employed/earner or not.(EARN)
(Yes=1)
0 278 47.93 47.93
1 302 52.07 100.00
Whether she has ownership of land or property or
not.(OWLP) (Yes=1) 0 86 14.83 14.83
1 494 85.17 100.00
Whether she has access to formal savings,
insurance, loan etc.(ASIL) (Yes=1) 0 262 45.17 45.17
1 318 54.83 100.00
Whether she has access education or training
service when she needs it.(AEDU) (Yes=1) 0 289 49.83 49.83
1 291 50.17 100.00
Whether anybody threats you to evict from
property (THPRO) (Yes=0) 0 24 4.14 4.14
1 556 95.86 100.00
Whether your present occupation is secured or not
(SEOC) (Yes=1) 0 281 48.45 48.45
1 299 51.55 100.00
Po
liti
cal
Dim
ensi
on
How does she involve in political process?
(INVOL) (Yes=1) 0 460 79.31 79.31
1 120 20.69 100.00
Whether she is a volunteer of any political party
(VOLP) (Yes=1) 0 533 91.90 91.90
1 47 8.10 100.00
Whether she attends any political gathering or not
(ATPM). (Yes=1) 0 331 57.07 57.07
1 249 42.93 100.00
Did you ever contest vote as a representative
(CTEST) (Yes=1) 0 559 96.38 96.38
1 21 3.62 100.00
Are you leader of any organization (LEAD)
(Yes=1) 0 381 65.69 65.69
1 199 34.31 100.00
So
cio
-Cu
ltu
ral
Dim
ensi
on
Whether she is a member of any social
organization or group (MEMSO) (Yes=1) 0 307 52.93 52.93
1 273 47.07 100.00
Whether she can influence the election/ selection
in the organization/group (INFLU) (Yes=1) 0 452 77.93 77.93
1 128 22.07 100.00
Does she participate in community activity?
(CACT) (Yes=1) 0 210 36.21 36.21
1 370 63.79 100.00
Whether or not she knows the location of the
nearest post-office, school, hospital, club,
vegetable market, etc. (KLSI) (Yes=1)
0 27 4.66 4.66
1 553 95.34 100.00
Does she feel exclusion from participation in any
community activity? (SEXL) (Yes=0) 0 86 14.83 14.83
1 494 85.17 100.00
Does she oppose the social curses a) Dowry
system, b) against Inter-caste marriage, c)
preference of male child d) against women
reservation e) child marriage f) against family
planning (OSOC)
(Value indicates number of opposed item )
0 9 1.55 1.55
1 78 13.45 15.00
2 203 35.00 50.00
3 139 23.97 73.97
4 113 19.48 93.45
5 31 5.34 98.79
6 7 1.21 100.00 Source: Author’s computation based on primary data, 2012-13
136
Table-4.3.2B Percentage Distribution of the Indicators of Women’s Empowerment
at the Community Level
Indicator Value Frequency Percent
Cumulative
Percent
Per
son
al/F
amil
ial
Dim
ensi
on
Did she ever campaign against social curse like
dowry, violence? (CASC) (Yes=1)
0 547 94.31 94.31
1 33 5.69 100.00
Whether she has professional training or not
(PTRN)(Yes=1)
0 421 72.59 72.59
1 159 27.41 100.00
Has she voluntarily changed her occupation
after marriage?(CHAM) (Yes=1)
0 295 50.86 50.86
1 285 49.14 100.00
Did she sacrifice employment or membership
of any organization due to familial ground?
(SACE)(Yes=0)
0 7 1.21 1.21
1 573 98.79 100.00
Whether she immunize her children in due
time.(EMUN) (Yes=1)
0 58 10.00 10.00
1 522 90.00 100.00
Leg
al D
imen
sio
n
Whether she ever used the local mechanism to
justices (ACCJ) (Yes=1)
0 538 92.76 92.76
1 42 7.24 100.00
Is she active in complaining about any problem
to the system of justice? (ACTC) (Yes=1)
0 189 32.59 32.59
1 391 67.41 100.00
Whether authorities are more or less effective
when other people’s concern compared to
yours? (AATY) (More= 0, less=1).
0 157 27.07 27.07
1 423 72.93 100.00
Whether she complains about the deficiency of
public services in her locality (CDPS) (Yes=1)
0 542 93.45 93.45
1 38 6.55 100.00
Source: Author’s computation based on primary data, 2012-13
Let us concentrate on table-4.3.2B where the indicators of personal/familial dimension
and legal dimension of women’s empowerment at community level have been presented.
It is revealed that only 5.69% of sample women have campaigned against social curses
like dowry, violence etc. This table shows that 27.41% have professional training and
49.14% have voluntarily changed their occupation after marriage. This table shows that
1.21% sacrifice their employment or membership of any organization due to familial
ground. We find from this table that 90% of sample women immunize their children in
due time. It is good for the health of future generation.
Statistics of the indicators of the legal dimension shows that 7.24% have got the help of
local mechanism to justice in contrast to 67.41% is active in complaining about any
problem to the system of justice. About 27.07% of respondents complain that the
authorities are more effective to other people compared to them in the matter of justice.
In our sample 6.55% women complain about the deficiency of public services in their
137
locality. The description of the indicators of the women’s empowerment at the
community level shows that in some cases women are advantageous and in other cases
they are disadvantageous in position. Therefore, it is not sufficient to study the overall
empowerment at the community level. We, therefore, have planned to construct
empowerment indices at the household level and at community level. The findings of the
empowerment indices have been discussed in the next section.
4.4. Analysis of Women’s Empowerment Indices
In order to quantify the empowerment level for the women in the district of Bankura we
have developed the women’s empowerment index in two ways. We have first computed
the degree of empowerment at the household level as well as at the community level.
The degree of empowerment has been computed by the ratio of the number of favourable
responses of the indicators of empowerment to the total number of indicators set for
capturing the empowerment. The degree of empowerment has been expressed as
percentage. This section along with its subsections gives the descriptive statistics of the
degree of empowerment and the descriptive statistics of the composite index of women’s
empowerment at the household level as well as at the community level.
4.4.1. Outcomes of Principal Component Analysis of the Household Level
Empowerment of Women
Before going to discuss the findings of the PCA we have checked the applicability of this
analysis in our study sample. For this purpose we have applied Kaiser-Meyer-Olkin
(KMO) measure of sample adequacy and Bartlett’s test of Sphericity. Now we look at
the findings of these tests.
Table 4.4.1 Results of KMO and Bartlett’s Test for Sample Adequacy for Factor
Analysis of Women’s Empowerment at the Household Level
Kaiser-Meyer-Olkin Measure of Sampling Adequacy 0.718
Bartlett's Test of Sphericity
Approx. Chi-Square 2384.175
Degrees of freedom 351
Significance. 0.000
Source: Author’s computation based on primary data, 2012-13
KMO measure evaluates the total correlation and partial correlations to determine
whether the data are likely to coalesce on components. By definition the value of KMO
138
measure ranges from 0 to 1. From the above table- 4.4.1 we find that the value of Kaiser-
Meyer-Olkin (KMO) measure of sampling adequacy is 0.71 (greater than 0.6 is desirable
as prescribed by UCLA Academic Technology Service). It indicates that the sample size
of this study is adequate for Principal Component Analysis. The Bartlett’s Test of
Sphericity examines whether the correlation matrix of the variables under consideration
is an identity matrix. The value of chi-square statistic in Bartlett’s Test of Sphericity is
also statistically significant. It confirms that the selected indicators of empowerment of
women are inter-correlated. Therefore, PCA is appropriate methodology for analyzing
the importance of the selected indicators in measuring empowerment of women at the
household level.
Table-4.4.2 presents the detailed results of the Principal Component Analysis which
reveal the variance explained by the components at initial situation, after extraction and
after rotation. The components have been extracted using Kaiser Criteria and rotated
imposing Varimax rotation method. This study has considered nine Principal
Components which have Eigen values greater than one. Rotation Sums of Squared
Loadings indicate that the first rotated principal component explains 10.46 per cent of
total variation in household level empowerment of women. The second and third rotated
principal components explain 7.41 per cent and 6.51 per cent of variance respectively.
Nine components altogether account for 55 per cent of total variation in the
empowerment of women at the household level.
Table-4.4.3 shows the rotated component matrix of women’s empowerment at the
household level. It represents rotated factor loadings. We see that the indicators
belonging to legal dimension have got highest importance in the construction of
Component-1. It indicates that legal dimension is the most important factor of women’s
empowerment at household level. Most of the important indicators of Components-2 and
Component-4 are belonging to socio-cultural dimension. Component-3 has been
weighted high by the indicators of economic dimension. The Component-5 and the
Component-8 capture the political dimension of the empowerment. The indicators that
load high on Component-6 and Component-9 are belonging to personal/familial
dimension. The indicators with highest loading in Component-7 don’t cover any specific
dimension. Therefore, the components under consideration cover all dimensions and are
not inter-correlated. Finally, weighted sum of the components, where weights are the
139
percentage of variance in the data set of the indicators explained by the particular
component, has been considered as the composite index of empowerment at the
household level for the women in the district of Bankura.
Table-4.4.2 Total Variance in the Indicators of Women’s Empowerment Explained
by the Components at the Household Level
Extraction Method: Principal Component Analysis
Rotation Method: Varimax
Co
mp
on
ent Initial Eigen values
Extraction Sums of Squared
Loadings
Rotation Sums of Squared
Loadings
Total % of
Variance
Cumulative
% Total
% of
Variance
Cumulative
% Total
% of
Variance
Cumulative
%
1 3.432 12.712 12.712 3.432 12.712 12.712 2.824 10.459 10.459
2 2.102 7.785 20.496 2.102 7.785 20.496 2.001 7.412 17.870
3 1.831 6.783 27.279 1.831 6.783 27.279 1.758 6.511 24.381
4 1.385 5.130 32.409 1.385 5.130 32.409 1.751 6.484 30.865
5 1.306 4.837 37.246 1.306 4.837 37.246 1.424 5.275 36.140
6 1.272 4.709 41.955 1.272 4.709 41.955 1.410 5.222 41.363
7 1.214 4.496 46.451 1.214 4.496 46.451 1.222 4.527 45.889
8 1.139 4.220 50.671 1.139 4.220 50.671 1.197 4.434 50.323
9 1.066 3.950 54.621 1.066 3.950 54.621 1.160 4.298 54.621
10 0.971 3.597 58.218
11 0.928 3.437 61.655
12 0.920 3.406 65.061
13 0.889 3.292 68.353
14 0.816 3.021 71.374
15 0.793 2.936 74.310
16 0.780 2.887 77.198
17 0.754 2.793 79.990
18 0.743 2.752 82.742
19 0.714 2.644 85.386
20 0.669 2.478 87.864
21 0.647 2.397 90.261
22 0.600 2.221 92.482
23 0.508 1.881 94.363
24 0.490 1.814 96.177
25 0.463 1.713 97.891
26 0.354 1.313 99.204
27 0.215 .796 100.00
Source: Author’s computation based on primary data, 2012-13
140
Table-4.4.3 Rotated Component Matrix of Women’s Empowerment at Household
Level
Extraction Method: Principal Component Analysis
Rotation Method: Varimax with Kaiser Normalization
Rotation converged in 9 iterations
Indicator Component
1 2 3 4 5 6 7 8 9
eco
no
mic
COPI 0.099 0.504 0.234 -0.024 -0.220 0.116 0.023 -0.241 -0.030
ACHR 0.001 0.054 0.599 0.027 -0.111 0.086 0.138 -0.083 0.001
PHESB -0.062 -0.140 0.585 -0.050 0.091 -0.056 0.187 0.458 0.101
DUSOL -0.003 0.085 0.717 -0.003 0.138 -0.026 -0.155 0.012 -0.006
SOBHA 0.019 0.112 0.486 0.232 -0.019 0.226 -0.066 -0.262 -0.014
CHOCU 0.054 -0.193 0.088 0.028 -0.029 -0.003 0.653 0.187 -0.028
po
liti
cal
DSPE 0.067 0.013 0.003 0.105 0.727 -0.033 0.049 -0.178 0.008
KLPL 0.004 0.632 -0.085 0.009 0.251 0.055 -0.257 0.090 -0.028
KOPL 0.173 0.156 0.047 0.164 0.698 0.104 -0.050 0.080 0.049
CASVO -0.001 0.264 -0.017 -0.299 0.360 0.112 0.476 0.011 -0.095
INVO 0.153 0.124 -0.127 0.092 -0.157 0.120 0.068 0.667 0.029
So
io-c
ult
ura
l
ENTM 0.219 0.646 -0.049 0.264 0.069 -0.094 -0.059 0.087 0.012
FMOH 0.327 0.480 0.130 -0.114 0.119 0.146 0.098 -0.158 0.053
GEDU -0.128 0.192 -0.196 0.283 -0.032 -0.069 0.473 -0.282 0.156
PLCP 0.067 0.576 0.133 0.071 0.036 -0.263 0.156 0.110 0.055
CLAB -0.066 0.074 0.084 0.768 0.127 -0.092 0.022 0.010 -0.027
CMAR 0.007 0.034 0.032 0.793 0.114 0.033 0.009 0.043 0.008
per
son
al/f
amil
ial
MATY -0.076 0.063 0.045 -0.097 0.065 0.157 0.085 0.184 0.704
ARTIP 0.044 0.178 -0.037 0.300 -0.084 0.282 0.114 -0.133 -0.432
INFE 0.067 0.034 0.171 0.115 -0.028 -0.556 0.056 -0.231 0.018
INDEC 0.045 0.029 0.174 0.012 -0.064 0.593 0.278 -0.011 0.004
PDEC 0.016 -0.070 0.177 0.027 0.118 0.635 -0.143 -0.063 0.103
leg
al
KAMJ 0.571 0.275 -0.181 -0.029 0.003 0.041 -0.142 0.085 0.006
TRMW 0.835 0.043 0.050 -0.031 0.055 0.049 0.006 0.024 -0.025
LAFW 0.879 0.056 0.000 -0.012 0.091 -0.062 0.059 0.004 -0.013
PUBS 0.863 0.097 0.052 0.031 0.071 -0.040 0.037 0.038 0.017
MARR 0.113 0.053 -0.042 0.231 -0.075 0.012 -0.052 -0.287 0.639
Source: Author’s computation based on primary data, 2012-13
141
4.4.2. Outcomes of Principal Component Analysis of the Community Level
Empowerment of Women
The value of Kaiser-Meyer-Olkin (KMO) measure of Sampling Adequacy for studying
women’s empowerment at community level in table-4.4.4 is 0.707 (greater than 0.6 is
desirable). It also ensures that the sample size in this study is adequate for Principal
Component Analysis. The value of chi-square statistic in Bartlett’s Test of Sphericity is
highly statistically significant. It ensures that the selected indicators of empowerment of
women at the community level are inter-correlated. Therefore, here PCA is also an
appropriate methodology for analyzing the importance of the selected indicators in
measuring empowerment of women at the community level. We have depicted the major
findings of PCA for women’s empowerment at the community level in table-4.4.5 and in
table-4.4.6.
Table-4.4.4 Results of KMO and Bartlett’s Test for Sample Adequacy for Factor
Analysis of Women’s Empowerment at the Community Level
Kaiser-Meyer-Olkin Measure of Sampling Adequacy 0.707
Bartlett's Test of Sphericity
Approx. Chi-Square 1989.652
Degrees of freedom 325
Significance 0.000
Source: Author’s computation based on primary data, 2012-13
Table-4.4.5 reveals the variance explained by the components at initial situation, after
extraction and after rotation. We have extracted the components applying Kaiser Criteria
and rotated imposing Varimax rotation method. We have found in table-4.4.5 that nine
principal components have Eigen values greater than one. So, this study has considered
nine Principal Components for measuring the composite index of women’s
empowerment at the community level. Rotation Sums of Squared Loadings indicate that
the first rotated principal component explains 9.29 per cent of total variation in
community level empowerment of women. The second and third rotated principal
components explain 7.47 per cent and 6.94 per cent of variance respectively. Nine
components altogether account for 55.26 per cent of total variation in the empowerment
of women at the community level.
142
Table-4.4.5 Total Variance in the Indicators of Women’s Empowerment Explained
by the Components at the Community Level
Extraction Method: Principal Component Analysis
Rotation Method: Varimax
Co
mp
on
ent
Initial Eigenvalues Extraction Sums of Squared
Loadings
Rotation Sums of Squared
Loadings
Total % of
Variance
Cumulative
% Total
% of
Variance
Cumulative
% Total
% of
Variance
Cumulative
%
1 3.223 12.396 12.396 3.223 12.396 12.396 2.416 9.292 9.292
2 2.048 7.876 20.272 2.048 7.876 20.272 1.944 7.478 16.769
3 1.617 6.219 26.491 1.617 6.219 26.491 1.806 6.946 23.715
4 1.561 6.003 32.494 1.561 6.003 32.494 1.672 6.430 30.145
5 1.465 5.633 38.127 1.465 5.633 38.127 1.595 6.136 36.282
6 1.304 5.014 43.141 1.304 5.014 43.141 1.396 5.368 41.650
7 1.086 4.177 47.318 1.086 4.177 47.318 1.283 4.935 46.585
8 1.046 4.024 51.342 1.046 4.024 51.342 1.174 4.516 51.100
9 1.020 3.921 55.263 1.020 3.921 55.263 1.082 4.163 55.263
10 .990 3.809 59.072
11 .964 3.709 62.781
12 .918 3.529 66.310
13 .907 3.490 69.800
14 .841 3.234 73.035
15 .769 2.958 75.993
16 .732 2.815 78.807
17 .688 2.647 81.455
18 .673 2.589 84.044
19 .627 2.413 86.457
20 .611 2.348 88.805
21 .592 2.276 91.080
22 .540 2.075 93.156
23 .477 1.834 94.990
24 .469 1.805 96.795
25 .444 1.708 98.502
26 .389 1.498 100.000
Source: Author’s computation based on primary data, 2012-13
143
Table-4.4.6 Rotated Component Matrix of Women’s Empowerment at the
Community Level
Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization.
Rotation converged in 9 iterations.
Indicator Component
1 2 3 4 5 6 7 8 9
eco
no
mic
EARN 0.213 -0.052 0.169 -0.436 -0.014 0.193 -0.330 0.301 0.036
OWLP 0.034 -0.277 0.355 0.160 0.179 0.280 0.115 0.346 -0.269
ASIL 0.736 0.006 0.179 0.169 -0.066 -0.004 0.064 -0.002 0.044
AEDU -0.017 0.084 0.000 0.562 0.180 -0.025 0.216 -0.003 0.212
THPRO -0.063 0.100 0.007 -0.234 0.028 0.091 0.175 0.612 0.214
SEOC -0.060 -0.104 -0.007 0.583 -0.087 0.071 0.142 -0.030 -0.054
po
liti
cal
INVOL 0.019 0.724 0.190 0.064 0.014 0.210 -0.024 -0.032 0.046
VOLP 0.023 0.701 0.009 0.095 0.021 0.067 -0.082 0.022 -0.040
ATPM 0.120 0.630 0.089 -0.164 0.085 0.147 0.216 0.007 -0.072
CTEST 0.180 0.397 -0.093 0.150 -0.041 -0.266 -0.088 0.487 0.025
LEAD 0.779 0.033 0.024 -0.040 0.017 0.173 0.050 -0.127 0.049
So
io-c
ult
ura
l
MEMSO 0.252 0.304 -0.071 -0.038 -0.128 0.606 0.056 -0.128 0.045
INFLU 0.135 0.217 0.034 0.178 0.035 0.676 -0.024 0.029 -0.023
CACT -0.042 -0.016 0.560 0.127 0.060 0.357 -0.225 0.069 0.297
KLSI 0.037 -0.070 0.025 0.066 0.028 0.013 0.031 0.009 0.849
SEXL 0.100 0.113 0.219 0.113 0.068 -0.117 0.606 -0.205 0.189
OSOC -0.009 0.117 -0.003 0.674 0.002 0.152 -0.187 0.143 0.003
per
son
al/f
amil
ial
CASC -0.002 0.112 0.106 0.125 0.700 -0.136 -0.089 -0.084 -0.064
PTRN 0.633 0.040 0.031 -0.181 -0.134 0.057 0.044 -0.006 -0.111
CHAM 0.772 0.116 0.058 -0.092 0.035 0.061 -0.139 0.016 0.061
SACE -0.133 -0.107 0.029 0.136 -0.025 -0.068 -0.086 0.448 -0.136
EMUN -0.028 -0.071 -0.128 0.096 -0.076 0.104 0.689 0.167 -0.088
leg
al
ACCJ -0.032 -0.025 -0.159 -0.113 0.697 0.081 -0.061 0.109 0.062
AATY 0.167 0.093 0.775 -0.184 -0.102 0.069 -0.022 -0.007 -0.024
ACTC 0.138 0.204 0.732 0.034 -0.073 -0.231 0.106 -0.010 -0.035
CDPS -0.083 -0.001 -0.043 0.020 0.681 0.016 0.105 -0.028 0.032
Source: Author’s computation based on primary data, 2012-13
In order to understand the relative importance of the indicators we have computed the
rotated component matrix of women’s empowerment at community level. Table-4.4.6
represents rotated factor loadings of the women’s empowerment at community level. We
144
see that the indicators that have more weight on Component-1 are; financial inclusion,
leadership, professional training and freedom of changing occupation. Therefore, these
are the most important factor of women’s empowerment at community level. They
altogether explain 10% of the total variation among the indicators of women’s
empowerment at the community level. All of the indicators that have high contribution
on Components-2 are belonging to political dimension. Component-3 and Component-5
have been weighted high by the indicators of legal dimension. The indicators of the
economic dimension load high on Component-4. The indicators that contribute high on
Component-6 and Component-9 are belonging to socio-cultural dimension. The
indicators with highest loading in Component-7 are social exclusion and child
immunisation. These are explaining 5% of total variance. The indicators namely whether
the woman faces evicting from own property, whether she contested any vote and
sacrifices employment or membership due to familial ground load high on Component-9.
Finally, similar to the composite index of empowerment at the household level we have
computed the composite index of empowerment at the community level for each woman
in the district of Bankura.
4.4.3. Descriptive Statistics of the Empowerment Indices
The Principal Component analysis has helped us construct the women’s empowerment at
the household level as well as at community level for each sample woman. We have got
two separate constructed variables for our sample women- composite index of women’s
empowerment at household level (CIWEH) and composite index of women’s
empowerment at community level (CIWEC). The descriptive statistics of these two
variables have been depicted in table-4.4.7. It is seen that the means of these variables
are close to zero. The median statistics of these two variables point out that the
empowerment at household level of the majority of the sample women is higher than its
mean while the empowerment at community level of the majority of the sample women
is lower than its mean. It says that majority of the women in the district of Bankura
enjoys higher empowerment in their households compared to in their community. The
Jarque Bera statistics for both the constructed variables, CIWEH and CIWEC, ensure
that the variables are normally distributed around mean zero. This is the main strength of
the PCA. This strength is reflected in this table-4.4.7. However, it is very difficult to
interpret the principal components and they have no unit of measurement. In order to
carry on with this shortcoming we have also constructed the degree of empowerment for
145
each sample women. We have seen that the average of the degree of empowerment at the
household level of the sample women is 68 in hundred points scale. This figure at the
community level is 54 in hundred points scale. Therefore, it also suggests that the
average empowerment of the women at the household level is higher than that at the
community level.
Table-4.4.7 Descriptive Statistics of the Women’s Empowerment
CIWEH CIWEC DOWEH DOWEC
Mean 3.00E-06 1.10E-06 68.21 54.38
Median 1.6 -0.24 70.37 53.85
Maximum 64.85 69.2 100 96.15
Minimum -52.6 -50.7 33.33 23.08
Std. Dev. 19.09 18.98 13.58 13.42
Skewness -0.19 0.17 -0.31 0.13
Kurtosis 2.83 3.16 2.69 2.76
Jarque-Bera 4.25 3.31 11.82 2.9
Probability 0.12 0.19 0 0.24
Source: Author’s own computation based on sample observations, 2012-13
Table-4.4.8 Percentage Distribution of Women’s Empowerment in Bankura
District
Levels of Empowerment As per Composite Index As per Simple Index
Number of Women Percent Number of Women Percent
Household Level Empowerment
Very low (-52≤CIWEH<-23) 72 12.41 65 11.21
Low (-23≤CIWEH<06) 271 46.72 211 36.38
Moderate (06≤CIWEH<35) 225 38.79 232 40.00
High (35≤CIWEH<65) 12 2.07 71 12.24
Community Level Empowerment
Very low (-50≤CIWEC<-20) 83 14.31 89 15.34
Low (-20≤CIWEC<10) 330 56.90 291 50.17
Moderate (10≤CIWEC<40) 156 26.90 182 31.38
High (40≤CIWEC<70) 11 1.90 18 3.10 Source: Author’s own computation, 2012-13
146
Figure-4.4.1 Percentage Distribution of Women’s Empowerment at Household
Level
Source: Author’s own computation based on table-4.4.8
Figure-4.4.2 Percentage Distribution of Women’s Empowerment at Community
Level
Source: Author’s own computation based on table-4.4.8
In addition to the mean median statistics, we have intended to analyse the relative level
of empowerment of the sample women shown in table-4.4.8. In accordance with the
range of the empowerment index values we have categorized the sample women into
four equal groups: high, moderate, low and very low. According to degree of
empowerment measure 12% (3%) of sample women have relatively high level of
empowerment at the household level (at the community level). About 40% (31%) of
147
sample women enjoy relatively moderate level of empowerment at the household level
(at the community level). We find that the women’s empowerment at the household level
(at the community level) is relatively low for 36% (50%) of the women. It is observed
that 11% (15%) have very low level empowerment at the household level (at the
community level). Therefore, we have found that a few women in the district of Bankura
have high level of empowerment. Majority of the women suffers from the lack of
empowerment at the household level and at the community level in the district of
Bankura.
With respect to the composite index computed applying PCA, we have seen that 2.07%
(1.9%) of sample women are highly empowered at the household level (at the
community level). On the other extreme, we observe that 12 % (14%) of the surveyed
women have very low level empowerment at the household level (at the community
level) and 39% (26%) of sample women enjoy relatively moderate level of
empowerment at the household level (at the community level). The empowerment at the
household level (at the community level) is relatively low for 46% (56%) of the women.
Therefore, the measure of composite index obtained following PCA, reveals more or less
same results regarding the level of empowerment of the rural women as we have
obtained from simple indices of empowerment.
4.5. Bivariate Correlation among Selected Variables
In this section, we test statistical significance of the correlation coefficients among the
different socio-economic and demographic variables of our sample women/households.
Although, the matrix of correlation coefficients considering all possible socio-economic
and demographic variables related to the surveyed population is computed, we discuss
only the correlation coefficients, which are relevant for our empirical models. First, we
analyze the strength of correlation between the household welfare indicating the decision
towards family planning, nature of domestic violence against women and expenditure for
education and women’s empowerment at the household level and at the community level
along with some individual/household and community characteristics of the surveyed
population. Second, we discuss the degree of association between the women’s
empowerment and the socio-economic and demographic traits, which have been
considered as determinants of empowerment of the rural women in the district of
Bankura. In order to test the significance of the correlation coefficient between two
148
particular variables we have used the test statistic 2
2,
1
r nt
r
with degrees of freedom
(n-2), where ‘r’ is the sample correlation coefficient between the pair of variables under
consideration and n stands for sample size. The results of the test of correlation
coefficients have been presented in the tables 4.5.1, 4.5.2, 4.5.3, 4.5.4and 4.5.5.
It has been revealed that women’s empowerment at the household level and at the
community level are highly correlated with the decision towards family planning. Refer
to table-4.5.1. The value of the correlation coefficient between the composite index of
empowerment at household level (community level) and family planning decision is 0.34
(0.47), which is statistically significant at 1% level. Thus, it is expected that the
probability of taking family planning increases with higher degree of empowerment of
women. The values of the bivariate correlation coefficients show that empowerment
variables are negatively correlated with the incidence of domestic violence against
women. These relations are statistically significant at 99% level of confidence. The
education expenditure as proportion to the annual household income has a positive and
statistically significant correlation with the women’s empowerment at the household
level and with the empowerment at the community level. Therefore, we can say that the
women’s empowerment reduces the incidence of domestic violence and inspire to spend
for education. The bivariate correlation analysis has justified our regression models for
assessing the impact of women’s empowerment on household welfare indicating the
decision towards family planning, nature of domestic violence against women and
expenditure for children’s education.
Table-4.5.1 Bivariate Correlation Matrix
DOWEH DOWEC CIWEH CIWEC DRFP PEDUX DVIO AGE
DOWEH 1 .605**
.947**
.530**
.324**
.126**
-.128**
-.039
DOWEC
1 .590**
.928**
.454**
.206**
-.116**
-.043
CIWEH
1 .503**
.345**
.124**
-.104* -.044
CIWEC
1 .477**
.199**
-.115**
-.069
DRFP
1 .079 -.056 -.094*
PEDUX
1 -.034 .060
DVIO
1 .000
AGE
1
*stands for significant at 5% level and ** stands for significant at 1% level
Source: Author’s own computation based on sample observations, 2012-13
149
Refer to table-4.5.2. It has been found that there is negative relation between the
women’s empowerment variables and the duration of married life. This relation is
significant only for composite index of empowerment at the household level. All
empowerment variables under consideration are highly correlated with the women’s
education and with the highest female education in the family. Women’s empowerments
at household level and at community level are negatively correlated with spousal age
gap. There is a negligible degree of association of the family planning decision with the
age at marriage and with the spousal age gap which are statistically insignificant. This
result points out that age at marriage and spousal age gap are immaterial to influence the
decision towards family planning. It is exhibited that there is a positive and significant
association between the education of the women and decision towards family planning.
Not only that, highest education level of male and female members in the family have
some favourable correlation with the decision towards family planning. It supports the
fact that educated couples usually prefer to have a small family.
Table-4.5.2 Bivariate Correlation Matrix
DURM AGAM HAGE SAGEG EDU HEDU EDUG HIMEDU HIFEDU
DOWEH -.062 .041 -.066 -.130**
.123**
.080 -.045 .035 .170**
DOWEC -.073 .055 -.060 -.144**
.235**
.222**
.028 .166**
.248**
CIWEH -.084* .075 -.077 -.148
** .126
** .077 -.054 .036 .140
**
CIWEC -.078 .014 -.067 -.088* .305
** .288
** .035 .222
** .299
**
DRFP -.110**
.027 -.075 -.023 .326**
.308**
.037 .262**
.206**
PEDUX .088* -.052 .041 .013 .093
* .127
** .074 .312
** .383
**
DVIO -.034 .065 -.016 -.011 -.261**
-.291**
-.101* -.260
** -.203
**
AGE .866**
.297**
.931**
.084* -.360
** -.259
** .089
* -.043 -.178
**
DURM 1 -.220**
.841**
.195**
-.361**
-.272**
.069 -.027 -.166**
AGAM
1 .211**
-.209**
-.013 .014 .041 -.034 -.029
HAGE
1 .349**
-.336**
-.239**
.087* -.023 -.180
**
SAGEG
1 -.037 -.022 .016 .029 -.018
EDU
1 .786**
-.139**
.543**
.712**
HEDU
1 .502**
.686**
.607**
EDUG
1 .340**
-.022
HIMEDU
1 .541**
HIFEDU
1
*stands for significant at 5% level and ** stands for significant at 1% level
Source: Author’s own computation based on sample observations, 2012-13
150
This table shows that women’s education and their husband education are significantly
and positively associated with their children’s education expenditure. The correlation
coefficient of education expenditure as a proportion of total income with mother’s
education and father’s education are 0.09 and 0.12 respectively. These results are
statistically significant at 5% level and 1% level respectively. Moreover, the highest
male and female education in the family positively affects the education expenditure in
the family. This result is also accepted at 99% confidence level. However, the degree of
domestic violence has insignificant association with duration of marriage, age at
marriage, husband age and spousal age gap. The correlation coefficient between
women’s education and degree of domestic violence is -0.26 which is statistically
significant at 1% level. Husband’s education is also negatively and significantly
associated with the degree of domestic violence. Besides highest male education, highest
female education and educational gap is negatively associated with the degree of
domestic violence. This result confirms that educational background of the family helps
reduce the intensity of domestic violence against women.
Refer to table-4.5.3. We see the correlation coefficient regarding the different types of
family occupation and other variables. There is a negative linear association between the
household level empowerment and the fact that the occupation of the family is
cultivation. We see that there is a positive and significant correlation coefficient between
the household level empowerment as well as community level empowerment and the
fact that the major occupation of the family is non-farm business. The correlation
coefficient between the community level empowerment and the wage-labour occupation
of the family is a negative and significant. There is a positive and significant correlation
coefficient between the household level empowerment as well as community level
empowerment and the fact that the earning members of the family are mainly employed
in the service sector. There is a negative and highly significant (1%) degree of linear
association between the household level empowerment and the dependency ratio in the
family. It has been identified that the association between the community level
empowerment and dependency ratio is insignificant. We see that the correlation
coefficient between the household level empowerment as well as community level
empowerment and average per capita household income is positive and significant at 1%
level. We find a positive and significant correlation between the community level
empowerment and household landholding.
151
Table-4.5.3 Bivariate Correlation Matrix
CULTI NFARM LABC SERV DRATIO APCHIN HLAND TYFAMI
DOWEH -.250**
.170**
-.008 .264**
-.114**
.179**
.036 .085*
DOWEC -.090* .156
** -.114
** .170
** -.058 .134
** .228
** .047
CIWEH -.274**
.186**
-.009 .290**
-.133**
.176**
.056 .027
CIWEC -.035 .173**
-.173**
.151**
.028 .158**
.249**
.033
DRFP -.010 .058 -.136**
.191**
-.006 .195**
.203**
-.043
PEDUX .048 .086* -.122
** .015 .372
** -.064 .060 .068
DVIO -.097* -.107
** .267
** -.157
** -.043 -.220
** -.071 -.007
AGE .040 .066 -.115**
.044 -.240**
.142**
.172**
.059
DURM .042 .075 -.102* .004 -.206
** .136
** .110
** .033
AGAM -.002 -.015 -.029 .079 -.076 .016 .126**
.051
HAGE .087* .041 -.133
** .021 -.215
** .131
** .201
** .028
SAGEG .139**
-.029 -.076 -.087* .053 -.009 .122
** -.071
EDU .100* .168
** -.370
** .259
** .251
** .304
** .245
** -.126
**
HEDU .153**
.193**
-.442**
.253**
.289**
.337**
.371**
-.134**
EDUG .106* .075 -.190
** .044 .112
** .115
** .253
** -.039
HIMEDU .184**
.206**
-.452**
.197**
.323**
.316**
.369**
-.142**
HIFEDU .126**
.163**
-.353**
.185**
.361**
.248**
.243**
-.096*
CULTI 1 -.365**
-.642**
-.249**
.210**
-.083* .362
** -.049
NFARM
1 -.294**
-.114**
.062 .045 .127**
.039
LABC
1 -.201**
-.252**
-.267**
-.499**
.046
SERV
1 -.025 .587**
.048 -.041
DRATIO
1 -.051 .104* -.089
*
APCHIN
1 .200**
-.012
HLAND
1 -.197**
TYFAMI
1
*stands for significant at 5% level and ** stands for significant at 1% level
Source: Author’s own computation based on sample observations, 2012-13
We get a positive and statistically significant correlation between simple index
household level empowerment and family composition. Therefore, family type may
determine the household level empowerment. The correlation coefficients of family
planning decision with annual per capita household income and with household
landholding are 0.195 and 0.203 respectively which are statistically significant.
Therefore, per capita family income and household landholding may be considered as
determinants of the probability of taking family planning decision. Each type of
household occupation is significantly correlated with the incidence of domestic violence.
The sign of correlation coefficient shows that women of labour class family are suffering
more violence compared to others. The correlation coefficient between the incidence of
152
violence and per capita family income is negative and significant. The figures of
correlation coefficients of education expenditure and households occupations tell us
households’ occupations affects education expenditure for children.
Table-4.5.4 Bivariate Correlation Matrix
GEN OBC SC ST LABOR FARM SELF HWIFE PERINC
DOWEH -.056 .060 .055 -.077 -.284**
.001 .066 .209**
.222**
DOWEC .045 .078 .011 -.180**
-.194**
.029 -.013 .151**
.126**
CIWEH -.076 .063 .066 -.070 -.275**
-.015 .083* .205
** .236
**
CIWEC .132**
.080 -.032 -.246**
-.179**
.050 -.033 .132**
.077
DRFP .088* .099
* -.057 -.167
** -.071 -.002 -.059 .093
* .068
PEDUX .003 .060 -.082* .041 -.013 -.005 .027 .002 -.026
DVIO -.18**
-.075 .177**
.094* .069 -.009 .026 -.065 -.115
**
AGE .005 .009 -.058 .071 -.026 -.032 -.005 .050 .064
DURM .013 .029 -.018 -.024 -.056 -.011 .018 .047 .012
AGAM -.014 -.037 -.079 .185**
.056 -.043 -.045 .009 .102*
HAGE .027 .002 -.058 .049 -.028 -.027 .002 .044 .033
SAGEG .094* -.051 -.055 .014 .007 -.015 .011 .000 -.087
*
EDU .299**
.121**
-.314**
-.136**
.009 .016 -.028 -.006 .149**
HEDU .349**
.122**
-.325**
-.195**
.005 -.004 -.015 .007 .109**
EDUG .141**
.027 -.082* -.122
** -.005 -.028 .015 .019 -.034
HIMEDU .244**
.131**
-.260**
-.149**
-.005 .011 .027 -.018 .075
HIFEDU .209**
.135**
-.282**
-.069 -.026 .040 -.005 -.007 .132**
CULTI .193**
.190**
-.282**
-.120**
.020 .002 .038 -.039 -.130**
NFARM .160**
.043 -.093* -.149
** -.036 .028 -.055 .037 -.034
LABC -.344**
-.242**
.404**
.224**
.041 .004 .000 -.039 -.028
SERV .042 .019 -.072 .022 -.065 -.050 .001 .095* .347
**
DRATIO .135**
.080 -.148**
-.084* .055 .033 .015 -.081 -.146
**
APCHIN .155**
.040 -.144**
-.064 -.065 -.029 .016 .070 .478**
HLAND .168**
.191**
-.255**
-.125**
-.046 -.027 -.001 .061 -.011
TYFAMI -.031 .005 .060 -.037 -.015 -.070 -.063 .100* -.039
GEN 1 -.371**
-.479**
-.245**
.061 -.041 -.041 .001 -.050
OBC 1 -.392**
-.201**
-.093* .015 -.025 .081 .076
SC
1 -.259**
-.046 .065 .060 -.043 -.060
ST
1 .093* -.053 .004 -.040 .063
LABOR
1 -.274**
-.156**
-.563**
-.064
FARM
1 -.133**
-.480**
-.043
SELF
1 -.274**
-.003
HWIFE
1 .090*
PERINC
1
*stands for significant at 5% level and ** stands for significant at 1% level
Source: Author’s own computation based on sample observations, 2012-13
153
In table-4.5.4 we see that the associations between household level empowerment and
caste variables are insignificant. But community level empowerment is significantly
related with dummies for GEN and ST. We have observed that the women who work for
wage have less empowerment compared to others. The correlation analysis reveals that
personal income of women is significantly correlated with the women’s empowerment at
the household level and at the community level. It has been seen that personal
occupation is significantly associated with women’s empowerment at household level as
well as at community level. In particular homemakers compared to others are enjoying
more empowerment at the household level and at the community level. Therefore, we
can adopt the personal income and personal occupation as determinants of women’s
empowerment at the household level and at the community level. Correlations of
personal income with the decision towards family planning and with the education
expenditure are insignificant. However, the personal income is negatively and
significantly correlated with the incidence of domestic violence.
Refer to table-4.5.5. We find a positive and statistically significant correlation of the
access to formal credit with the empowerment variables. This result suggests for
considering the access to formal credit as a determinant of women’s empowerment at the
household level and at the community level. There is a positive and significant
correlation between the decision towards family planning and access to formal credit and
between accessibility to formal credit and education expenditure as proportion to
household income. It indicates that access to formal credit is an influential factor in the
determination of the family planning decision and educational expenditure.
The bivariate correlation matrix shows that the degree of association between the decision
towards family planning and the duration of the SHG membership status for the women
under study is positive and significant. Further, we have found that the correlation
coefficient between the education expenditure as proportion to the annual household
income and the duration of SHG membership is positive and significant. However,
surprisingly we find a positive and statistically significant correlation between the
incidence of domestic violence and the duration of SHG membership status. Therefore, in
the area under study family co-members do not want that women participate in the SHG-
centric microfinance program in which women move outside home and interact with the
officials of the programs and her co-members.
154
Table-4.5.5 Bivariate Correlation Matrix
ALOAN DSHGM ADDIC DOW PMDOW
DOWEH .286**
.175**
.094* .033 .036
DOWEC .487**
.389**
.064 .130**
.026
CIWEH .277**
.202**
.143**
.004 .050
CIWEC .489**
.329**
.023 .207**
.028
DRFP .243**
.162**
.020 .092* .049
PEDUX .104* .146
** -.069 .069 -.022
DVIO -.009 .131**
.257**
.097* .236
**
AGE .031 .151**
.017 -.253**
-.098*
DURM .015 .083* -.053 -.210
** -.076
AGAM .032 .136**
.135**
-.091* -.048
HAGE .020 .131**
.002 -.247**
-.087*
SAGEG -.088* -.036 -.104
* -.090
* -.009
EDU .041 -.106* -.228
** .188
** -.064
HEDU .075 -.073 -.240**
.227**
-.097*
EDUG .063 .032 -.067 .100* -.067
HIMEDU .090* .009 -.182
** .184
** -.052
HIFEDU .078 -.024 -.194**
.164**
-.092*
CULTI -.048 -.079 -.210**
.175**
-.108**
NFARM .074 .038 -.002 .102* .013
LABC -.019 .083* .213
** -.212
** .139
**
SERV .026 -.051 .015 -.087* -.066
DRATIO -.015 -.065 -.135**
.172**
.038
APCHIN .066 -.131**
-.093* -.041 -.060
HLAND .163**
.172**
-.089* .122
** .004
TYFAMI .032 -.013 .059 .035 -.031
GEN .006 -.083* -.111
** .184
** -.117
**
OBC .049 -.029 -.146**
.156**
.032
SC .015 .026 .139**
.033 .165**
ST -.100* .125
** .153
** -.522
** -.113
**
LABOR -.132**
-.070 .028 -.004 .065
FARM .103* .032 .015 .019 .009
SELF -.034 -.056 -.057 .010 -.008
HWIFE .049 .064 -.006 -.016 -.059
PERINC .064 -.044 -.014 -.172**
-.056
AFCT 1 .542**
.084* .046 .068
DSHGM 1 .196**
-.079 .032
ADDIC
1 -.094* .062
DOW
1 .216**
PMDOW
1
*stands for significant at 5% level and ** stands for significant at 1% level
Source: Author’s own computation based on sample observation, 2012-13
155
The correlation coefficient between drug addiction by husband and incidence of domestic
violence is positive and statistically significant. This finding tells us that addiction may
induce the risk of domestic violence. It also supports the addiction as a determinant of
domestic violence. This study shows that dowry at marriage and the post marriage
demand for dowry are positively and significantly associated with the violence against
women. With this end in view, we have considered dowry at marriage and the post
marriage demand for dowry as important determinants of domestic violence in our
regression model.
The analysis of correlation coefficient among different socio-economic and demographic
traits of the women/households actually guides us to select the important explanatory
variables for the econometric models specified for analyzing the impact of women’s
empowerment on household welfare and for examining the influential factors of
empowerment for the women in the district of Bankura. The above explanation helps us
recognize the relevant explanatory variables regarding the models of the decision towards
family planning, incidence of domestic violence, spending for education and women
empowerment.
Once we have diagnosed and identified the explanatory variables/determinants for the
models under consideration, we shall be able to measure the impact of these
determinants. This exercise will come to our help in prescribing the economic policies in
connection with the women’s empowerment.
4.6. Conclusion
This chapter has presented the detailed descriptions of the sample women and their
households. A description of the indicators of empowerment shows that women in the
district of Bankura are highly deprived of several aspects of empowerment. We have
computed the degrees of household level as well as community level empowerment for
each sample woman. Using PCA a composite index of empowerment at the household
level and at the community level for each sample women has also been estimated. The
measures of the degree of empowerment and PCA constructed four empowerment
variables reveal that average levels of empowerment at the household level and at the
community level of the sample women are not commendable. The bivariate correlation
analyses suggests that empowerment at the household level as well as at the community
156
level are significantly associated with the decision regarding family planning, incidence
of domestic violence and with the spending for child education. Based on these
associations we have estimated the regression models to assess the impact of
empowerment variables on household and child welfare. We also examine the factors
affecting empowerment at the household level and at the community level. The results of
the empirical models have been discussed in chapter five.
157
Chapter Five
__________________________________________________
Empirical Estimates and Analysis
5.1. Introduction
We have formulated the relevant working models and hypotheses related to the study of
women’s empowerment in the district of Bankura. The descriptive statistics of several
household and individual characteristics including focussed constructed variables degree
of women’s empowerment and the composite index of women’s empowerment have
been explained in the last chapter. In this chapter our plan is to describe the empirical
findings and analyze the econometric models duly estimated.
The route of journey of this chapter has been designed as follows. In section 5.2 we have
interpreted and analyzed the estimated results of the model for the decision towards
family planning. The estimated impacts of women’s empowerment status along with the
other determinants on the incidence of domestic physical violence have been analyzed in
section 5.3. Section 5.4 has explained the impact of women’s empowerment along with
other socio-economic and demographic traits on the expenditure for children education.
Section 5.5 deals with the determinants affecting the women’s empowerment at
household level as well as at community level. The determinants of the household level
empowerment have been discussed in sub-section 5.5.1. Sub-section 5.5.2 has
interpreted the results of the estimates of community level empowerment of women in
the district of Bankura. Finally, section 5.6 concludes the chapter.
5.2. Impact of Women’s Empowerment on Decision regarding Family Planning
In chapter three we have specified an econometric model regarding the decision towards
family planning. It has been pointed out that we use a Probit model for assessing the
probability of taking family planning decision for the women in the district of Bankura.
158
The probability of taking family planning decision depends on a number of determinants
divided into individual/household and community characteristics of the women. Among
them most important variable is definitely the women’s empowerment. We expect that
after controlling for individual/household and community characteristics, the women
with higher level of empowerment at the household level and at the community level
have better position to take family planning decision. This implies that the likelihood of
taking family planning is expected to increase with higher empowerment at the
household level and at the community level. In addition to women’s empowerment at the
household level and at the community level, male child bias, age, spousal age gap,
education of the woman, husband education, household occupation, income,
landholding, dependency ratio and the duration of SHG-membership have been
considered as important explanatory variables in the probit model for the decision
regarding family planning.
Based on the methodology of computing empowerment we have estimated two models
for each issue of household welfare achieved through empowerment. In order to explain
the decision towards family planning we have estimated two models, Model-1A and
Model-1B, in each model all the variables being the same, except empowerment
variables. In Model-1A we consider the empowerment variables measured by simple
average method and Model-1B incorporates the composite index for the women’s
empowerment computed by principal component analysis. For each model the number of
observation is 580.
5.2.1. Model-1A: Probit Model with Simple Empowerment Indices
In this sub-section we discuss the results of the estimated probit model of the decision
regarding family planning where empowerment variables have been measured by simple
average of the indicators which is called the degree of empowerment, i.e., Model-1A.
The results of the Model-1A are presented in table-5.2.1 and in table-5.2.2. First, we
interpret the coefficients of the individual/household characteristics as explanatory
variable; then we come to the community characteristics.
In table-5.2.1 the coefficient of women’s empowerment at the household level is positive
and statistically significant. One percent higher household level empowerment increases
the log-odds in favour of taking family planning by 1.3 percentage points. It is indicative
159
that household level empowerment is an important factor for taking decision regarding
family planning. Refer to table-5.2.2. We have found that the probability of taking
decision regarding family planning increases by 0.5 percent point due to one percent
increase in the degree of empowerment from the mean level household empowerment of
the women. It implies that household level empowerment has a positive marginal impact
on the probability of taking family planning decision. Empowered women at the
household have the decision making power and consciousness regarding family and
child welfare. So, the result is consistent with the logic.
The coefficients of the dummies for first two child combinations turn out to be negative.
The dummy for first child male and second child female (FMSF=1) is statistically
insignificant to determine family planning decision. The dummy for first child female
and second male (FFSM=1) and the dummy for first and second children female
(FFSF=1) are statistically significant determinants of the decision regarding family
planning. The estimated probit model shows that the probability of taking family
planning decision is 26% lower for the women having first two children female than the
women having one child or first two children male. We also find that the women having
first child female and second child male are 13% less likely to adopt family planning in
contrast to the women having one child or first two children male. Therefore, women
having first two children female are worst in position to have family planning decision.
These findings prove that our sample women have male child bias and this is the reason
they did not want to take family planning before having two male children. So our male
child bias hypothesis regarding family planning turns out to be true.
It is revealed that age at marriage has some positive impact on the log-odds towards
family planning for the women in the district of Bankura. The marginal change of
probability states that if average age of the women at marriage increases by one year
probability of taking family planning decision increases by 0.6 percentage point from
mean. This result is, however, statistically insignificant. High age at marriage increases
the consciousness of the women regarding familial and biological know how of family
planning. So, the result is expected. We have also found that the coefficient of spousal
age gap is positive and statistically significant at 6% level. The magnitude of the
coefficient of spousal age gap says that age gap is suitable for positive decision
regarding family planning. We have seen that average age gap of our sample couples
160
(husband’s age – wife’s age) is 5.7 years. The marginal change of probability calculation
in table-5.2.2 implies that ceteris paribus one year increase in spousal age gap from mean
increases the probability of taking family planning by 1.6 % points. This result supports
the Indian culture regarding age of husband and wife and goes against our hypothesis.
Therefore, Indian culture may be suitable for taking family planning. However, the result
is not highly significant at all.
Table-5.2.1 Results of the Probit Model for the Decision regarding Family Planning
When Women’s Empowerment is the Simple Average of the Indicators
Dependent Variable: DRFP (Decision regarding Family Planning)
Method: ML - Binary Probit (Newton-Raphson)
Included observations: 580
Convergence achieved after 5 iterations
Covariance matrix computed using second derivatives
Explanatory Variable Coefficient Std.
Error
z-
Statistic Prob.
Constant -3.341 0.569 -5.875 0.000
Personal/Household characteristics
DOWEH (%) 0.013**
0.006 2.246 0.025
Women having at least two child (FMSF =1) -0.080 0.166 -0.480 0.631
Women having at least two child (FFSM =1) -0.355**
0.166 -2.134 0.033
Women having at least two child (FFSF =1) -0.695*
0.184 -3.780 0.000
Age at Marriage AGAM (Year) 0.016 0.013 1.194 0.233
Spousal Age Gap (SAGEG) (Year) 0.042***
0.023 1.827 0.068
Education of the Woman (EDU) (Year) 0.082*
0.026 3.087 0.002
Husband’s Education (HEDU) (Year) 0.020 0.024 0.870 0.384
Type of Family (TYFAMI) (Nuclear = 1) -0.258 0.169 -1.527 0.111
Household Occupation CULTI, (Cultivation =1) -0.144 0.173 -0.835 0.404
Household Occupation, NONFARM, (Non-Farm= 1) -0.437**
0.218 -2.009 0.045
Household’s Land holding, HLAND, (bigha, = 0.4hector) 0.014 0.029 0.474 0.636
Dependency Ratio in the Household DRATIO (%) -0.002 0.003 -0.589 0.556
Annual Per Capita Household Income (APCHIN) (Rs. ‘000) 0.023**
0.011 2.186 0.029
Community characteristics
DOWEC (%) 0.038*
0.007 5.304 0.000
Duration of SHG-membership DSHGM (Month) 0.003 0.002 1.530 0.126
Caste (OBC=1) 0.153 0.177 0.864 0.388
Caste (SC=1) 0.063 0.180 0.351 0.726
Caste (ST=1) -0.436***
0.243 -1.792 0.073
Summary Statistics
LR statistic (19 df) (probability) 224.996 (0.000) Akaike information Criterion 1.032
McFadden R-squared 0.287 Schwarz criterion 1.183 Source: Author’s own computation based on sample observations, 2012-13
*, ** and *** imply that coefficients are significant at level 1%, 5% and 10% respectively.
161
Table-5.2.2 Marginal Probability of the Decision regarding Family Planning When
Women’s Empowerment is the Simple Average of the Indicators
Marginal effects after probit
Dependent Variable(y): Probability of the Decision regarding Family Planning (DRFP)
(predict) = .6445
Included observations: 580
Explanatory Variable
dy/dx
evaluated
at mean
Std.
Error z-Statistic
Prob.
>|z|
Mean of the
Explanatory
Variable(x)
Personal/Household characteristics
DOWEH (%) 0.005**
.0022 2.25 0.025 68.12
Women having at least two child (FMSF=1)# -0.029 .0629 -0.48 0.634 0.20
Women having at least two child (FFSM =1) # -0.136
** .0649 -2.10 0.036 0.19
Women having at least two child (FFSF =1) # -0.269
* .0706 -3.81 0.000 0.16
Age at Marriage AGAM (Year) 0.006 .005 1.20 0.232 18.75
Spousal Age Gap (SAGEG) (Year) .0158***
.0086 1.83 0.068 5.74
Education of the Woman (EDU) (Year) 0.030* .0098 3.09 0.002 3.59
Husband’s Education (HEDU) (Year) 0.007 0.87 .0087 0.384 4.86
Type of Family (TYFAMI) (Nuclear = 1) # -0.092 .0582 -1.59 0.111 .818
Household Occupation CULTI, (Cultivation =1) #
-0.053 .0645 -0.84 0.404 0.443
Household Occupation NONFARM (Non-Farm= 1) # -0.169
** .0855 -1.98 0.048 0.143
Household’s Land holding, HLAND, (bigha) .0051 .0109 0.47 0.636 2.64
Dependency Ratio in the Household DRATIO (%) -.0006 .0011 -0.59 0.556 49.68
Per Capita Household Income (APCHIN) (Rs. ‘000) .0086**
.0038 2.22 0.027 13.78
Community characteristics
DOWEC (%) 0.014* .0026 5.31 0.000 54.37
Duration of SHG-membership DSHGM (Month) 0.001 .0007 1.53 0.126 27.23
Caste (OBC=1) # 0.056 0.88 0.06 0.378 0.23
Caste (SC=1) # 0.023 .066 0.35 0.725 0.33
Caste (ST=1) # -0.169
*** .096 -1.76 0.079 0.11
(#) dy/dx is for discrete change of dummy variable from 0 to 1
*, ** and *** imply that coefficients are significant at level 1%, 5% and 10% respectively. Source: Author’s own computation based on sample observations, 2012-13
Education of the women has also some direct impact on the decision regarding family
planning. Our marginal probability computation has shown that one additional
completed year of formal education after primary level increases the probability of
adopting family planning by 3 percentage points. Education of women inculcates the
knowledge and importance of family planning for family welfare as well as for social
welfare. Therefore, education is very much important factor affecting the decision
regarding family planning as we have found. The coefficient of husband’s education in
Model-1A is also positive but statistically insignificant. One year extra above primary
162
level (mean education of the husband) education of husband increases the probability of
adopting family planning by 0.8% point. By the same logic as we have mentioned for
women’s education the impact of husband’s education on the probability of adopting
family planning is consistent. However, impact of women’s education is more important
than that of husband’s education on the probability of adopting family planning.
Estimates of the probit model shows that the dummy for family type (Nuclear =1) has
negative effect on the log-odds in favour of family planning. It tells us that women
belonging to nuclear family are less likely to be ideal family planner. The probability
towards family planning reduces by 9.2 percentage point if the woman is belonging to
nuclear family. This result is statistically significant at 11% level.
We have examined the impact of household occupation on the attitudes of sample
women towards family planning. Household occupation, CULTI (cultivation =1) has
some negative impact on log-odds towards the decision regarding family planning. It
means that women belonging to cultivator family relative to women belonging to wage
labour class are less likely to adopt family planning. However, this empirical result is
statistically insignificant. The coefficient of household occupation, NONFARM (Non-
farm=1) in Model-1A points out that women in non-farm self-employed households
compared to wage labour class are less likely to take family planning. Table-5.2.2 shows
that, if a household shifts to self-employed occupation from service or wage earning
jobs, the probability of taking family planning will reduce by sixteen percentage points.
It may happen due to the fact that wage labour households/couples think the opportunity
cost of child rearing whereas women in farm and self-employed family do not think
regarding opportunity cost of child rearing. Rather in farm and petty business family
children are viewed as earner at least in rural area.
It turns out that dependency ratio in the households has some adverse but insignificant
impact on the decision regarding family planning. We find that the coefficient of
household’s landholding is positive in our estimated probit model. It means that
household’s landholding is favourable for adopting family planning decision but this
empirical finding is statistically insignificant. The coefficient of annual per capita
household income, which is significant at 2% level, reports that one thousand rupees
additional annual per capita household income increases the log-odds in favour of
adopting family planning by 2.3% points. Table-5.2.2 shows that if annual per capita
163
household income increases one thousand rupees probability of adopting family planning
would increase by almost 1% point. It is fact that rich households are very much
concerned regarding child care i.e. about health, education and future economic status
and thereby cost of child care. So, it is an expected result.
The prime community trait captured by the women’s empowerment at the community
level has some positive and statistically significant impact on the probability of adopting
family planning. As women’s empowerment increases 1% the probability of taking
family planning increases by 1.4% points. Not only that, community level empowerment
is more important than the household level empowerment in the determination of the
probability of adopting family planning. Therefore, employment outside home,
association with community affairs, and mobility of the women are most important for
taking family planning. Mobility of the women in different social institutions helps her
understand the importance of family planning.
The coefficient of the duration of SHG-membership is positive and statistically
significant at 12% level. It has been observed that one year extra participation in SHG
from mean increases the probability of taking family planning by 0.1%. Participation in
SHG ensures the financial inclusion of the women. As duration of participation increases
the intensity of financial inclusion increases ensuring higher amount of borrowing if
needed. A large number of sample women has reported that availing SHG loan they are
now self-employed. Therefore, availability of fund to the women increases the
opportunity cost of time to her. Besides, participation in SHG inculcates the
consciousness in women member. It induces them to take family planning. However, the
impact of SHG-membership is too small in the determination of the probability of taking
family planning.
In order to answer whether caste is a matter or not to take family planning decision we
have considered three dummies for caste of the women. In this study Caste (general caste
=1) has been considered as comparison category. The coefficients of dummies for caste
(OBC=1) and for caste (SC =1) indicate that the women belonging to other backward
classes and scheduled castes in contrast to general caste women are more likely to take
family planning decision. But these findings have no statistical base. We also find that
the coefficient of the dummy for caste (ST =1) is negative but significant at 7% level. So
164
the probability of adopting family planning for the tribal couple is 16% lower than that
of the general caste couples. It may arise due to ignorance, unawareness and lack of
consciousness of the tribal women.
This study reveals that empowerment of women at the household level and at the
community level, male child bias, education of woman, family type, household income,
participation in SHG and caste are significant determinants of the decision towards
family planning for the women in Bankura district.
5.2.2. Model-1B: Probit Model with Composite Empowerment Indices
In this section we consider the probit model (Model-1B) for the decision regarding
family planning where empowerment variables are composite indices along with the
same other explanatory variables as had in Model-1A. This new model is denoted by
Model-1B and results have been depicted in table-5.2.3 and table-5.2.4.
Refer to table-5.2.3 and table-5.2.4. The coefficients of the composite index of women’s
empowerment at household level and at community level are positive and statistically
significant at 1% level. Women’s empowerment at household level and at community
level increases the probability of taking family planning measures for the sample
women. It is important to note that compared to household level empowerment,
community level empowerment has higher impact on family planning decision.
Therefore, methodological difference in measuring women’s empowerment does not
alter the findings regarding the impact of women’s empowerment on the probability of
adopting family planning decision.
In addition to empowerment, male child bias has same line impact on the decision
regarding family planning as we have in Model-1A. Like Model-1A we have found that
women’s education is significant determinant of the decision towards family planning.
The estimates of the probit model show that one year extra education of the woman from
mean education increases the probability of taking family planning decision by 2.7
percent point while one year extra education of husband from mean education increases
the probability of taking family planning decision by 0.97 percent point. Therefore,
women’s education compared to husband education is more important in family planning
decision as we have seen in Model-1A.
165
Table-5.2.3 Results of the Probit Model for the Decision regarding Family Planning
When Women’s Empowerment is Composite Index of the Indicators
Dependent Variable: Decision regarding Family Planning (DRFP)
Method: ML - Binary Probit (Newton-Raphson)
Included observations: 580
Convergence achieved after 5 iterations
Covariance matrix computed using second derivatives
Explanatory Variable Coefficient Std. Error z-Statistic Prob.
Constant -0.356 0.423 -0.841 0.400
Personal/Household Characteristics
CIWEH 0.016*
0.004 3.758 0.000
Women having at least two child (FMSF =1) -0.059 0.168 -0.350 0.727
Women having at least two child (FFSM =1) -0.367**
0.169 -2.174 0.030
Women having at least two child (FFSF=1) -0.708*
0.187 -3.795 0.000
Age at Marriage AGAM (Year) 0.015 0.014 1.144 0.253
Spousal Age Gap (SAGEG) (Year) 0.041***
0.024 1.752 0.080
Education of the Woman (EDU) (Year) 0.073*
0.027 2.719 0.007
Husband’s Education (HEDU) (Year) 0.026 0.024 1.098 0.272
Type of Family (TYFAMI) (Nuclear = 1) -0.230 0.171 -1.345 0.179
Household Occupation CULTI, (Cultivation =1)
-0.123 0.175 -0.700 0.484
Household Occupation NONFARM (Non-Farm= 1) -0.490**
0.221 -2.219 0.027
Household’s Land holding, HLAND, (bigha) 0.014 0.030 0.471 0.637
Dependency Ratio in the Household DRATIO (%) -0.002 0.003 -0.770 0.441
Per Capita Household Income (APCHIN) (Rs. ‘000) 0.021**
0.010 2.046 0.041
Community Characteristics
CIWEC 0.026*
0.005 5.492 0.000
Duration of SHG-membership DSHGM (Month) 0.003 0.002 1.561 0.119
Caste (OBC=1) 0.161 0.180 0.894 0.372
Caste (SC=1) 0.093 0.183 0.508 0.611
Caste (ST=1) -0.329 0.250 -1.317 0.188
Summary Statistics
LR statistic (19 d f) (Probability) 238.393 (0.000) Akaike information Criterion 1.009
McFadden R-squared 0.304 Schwarz Criterion 1.160 Source: Author’s own computation based on sample observations, 2012-13
*, ** and *** imply that coefficients are significant at level 1%, 5% and 10% respectively.
The coefficient of household occupation (Non-farm=1) in Model-1B points out that non-
farm self-employed households compared to wage labour class are less likely to take
family planning. Table-5.2.4 shows that if a household shifts to self-employed
occupation from service or wage earning jobs the probability of taking family planning
will reduce by 18 percentage points. Like Model-1A we find that household income
increases the probability of taking family planning as we expect. The coefficient of the
duration of SHG-membership is positive and statistically significant at 11% level. One
166
year extra participation in SHG from mean increases the probability of taking family
planning by 1.1% magnifying the consciousness in the member women that induces
them to take family planning. However, age at marriage, spousal age gap, family
composition, household’s landholding and caste are insignificant factors determining the
probability of accepting family planning decision.
Table-5.2.4 Marginal Probability of the Decision regarding Family Planning When
Women’s Empowerment is measured by Composite Index
Marginal effects after probit
Dependent Variable(y): Probability of the Decision regarding Family Planning (DRFP)
(predict) = .6457
Included observations: 580
Explanatory Variable
dy/dx
evaluated
at mean
Std. Error z-Statistic Prob.
>|z|
Mean of the
Explanatory
Variable(x)
Personal/Household Characteristics
CIWEH 0.005* .0015 3.77 0 3.00E-06
Women having at least two child (FMSF =1)# -0.021 .0631 -0.35 0.728 0.208
Women having at least two child (FFSM=1) # -0.140
** .0659 -2.14 0.033 0.191
Women having at least two child (FFSF =1) # -0.274
* .0716 -3.83 0 0.163
Age at Marriage AGAM (Year) 0.005 .0050 1.15 0.252 18.75
Spousal Age Gap (SAGEG) (Year) 0.015***
.0087 1.75 0.08 5.743
Education of the Woman (EDU) (Year) 0.027* .01 2.73 0.006 3.593
Husband’s Education (HEDU) (Year) 0.009 .0089 1.1 0.272 4.867
Type of Family (TYFAMI) (Nuclear = 1) # -0.082 .0593 -1.4 0.162 0.818
Household Occupation CULTI, (Cultivation =1) #
-0.045 .0654 -0.7 0.484 0.443
Household Occupation NONFARM (Non-Farm= 1) # -0.189
** .0866 -2.19 0.029 0.143
Household’s Land holding, HLAND, (bigha) 0.005 .011 0.47 0.637 2.646
Dependency Ratio in the Household DRATIO (%) -0.001 .0011 -0.77 0.441 49.68
Per Capita Household Income (APCHIN) (Rs. ‘000) 0.007**
.0038 2.07 0.038 13.78
Community Characteristics
CIWEC 0.009* .0017 5.48 0 1.10E-06
Duration of SHG-membership DSHGM (Month) 0.001 .0007 1.56 0.118 27.23
Caste (OBC=1) # 0.058 .064 0.91 0.362 0.232
Caste (SC=1) # 0.034 .0672 0.51 0.609 0.336
Caste (ST=1) # -0.126 .0983 -1.29 0.198 0.117
(#) dy/dx is for discrete change of dummy variable from 0 to 1 *, ** and *** imply that coefficients are significant at level 1%, 5% and 10% respectively.
Source: Author’s own computation based on sample observations, 2012-13
We, therefore, observe that almost all the findings remain intact in Model-1B as they
were in Model-1A. The major findings in this section are as follows. Women’s
167
empowerments at the household level as well as at the community level are significant
determinants of the decision regarding family planning. We find that male child bias,
education of women, non-farm household occupation, dependency ratio, per capita
income and participation in SHG are important factors influencing the decision regarding
family planning.
5.3. Impact of Women’s Empowerment on Domestic Violence against Women
In this section we have explored the determinants of the incidence of domestic violence
for the women in the district of Bankura. At the outset, we would look into the nature of
domestic violence against women with respect to their age. We have observed that
younger and older women are experiencing higher risk of violence relative to the middle
aged group women. It indicates that newly married women (mean age at marriage is 19
years) and older suffers from more violence than the others. This finding supports that in
the middle age when women are physically and mentally more active to protest violence
and other family members become scared to make violence against women. However,
figure-5.3.1 shows that in all age group a major section of women is suffering from
domestic violence.
Figure-5.3.1 Age-Group Wise Prevalence of Domestic Violence against Women in
Bankura District
Source: Author’s own computation based on sample observations, 2012-13
In figure-5.3.1 we see that the relation between age and the incidence of domestic
violence against women is non linear. If we try to draw the line joining the mid points of
168
the bars we get an approximately U shaped curve. It is already revealed in the profile of
inter correlation (refer to table-4.5.1) that the correlation between age and the incidence
of domestic violence is almost zero. We could not include age of the women as
determinant of the probability of the incidence of domestic violence, although, most of
the existing empirical study regarding domestic violence has included the age as
determinant of domestic violence.
So far, the examination of the impact of women’s empowerment on the probability of
experiencing domestic violence is our main motto. In addition to women’s
empowerment we have included spousal age gap, husband education and education of
other family members, household occupation, income, dependency ratio, dowry, drug
addiction of the husband, participation in SHG and caste of the women/household as
determinants of the incidence of domestic violence.
In accordance with the methodology of measuring women’s empowerment we have
estimated two models Model-2A and Model-2B for the incidence of domestic violence.
In Model-2A we consider the empowerment variables measured by simple average
method and Model-2B incorporates the composite indices of women’s empowerment
computed by PCA. In each model we have 580 observations.
5.3.1. Model 2A: Logit Model with Simple Empowerment Indices
In this sub-section, we discuss the results of the estimated logit model of the incidence of
domestic violence where empowerment variables have been measured by simple average
of the indicators. This average is termed as the degree of empowerment. The outcomes
of the Model-2A have been presented in table-5.3.1 and in table-5.3.2.
The coefficient of empowerment of women at the household level is -0.024, which is
statistically significant at 1.2% level. It indicates that other things remaining the same,
the household level empowerment of women reduce the log odds towards domestic
violence against them. The marginal change of probability has reported that one percent
increase in women’s empowerment at the household level from mean reduces the
probability of the incidence of domestic violence by 0.6% point. An empowered woman
can logically establish her views and she has some decision making power.
169
Table-5.3.1 Results of the Logit Model for the Incidence of Domestic Violence
When Women’s Empowerment is Simple Average of the Indicators
Dependent Variable: Incidence of Domestic Violence (DVIO)
Method: ML - Binary Logit (Newton-Raphson)
Included observations: 580
Convergence achieved after 4 iterations
Covariance matrix computed using second derivatives
Explanatory Variable Coefficient Std. Error z-statistic Prob.
Constant 1.650 0.936 1.764 0.078
Individual/household Characteristics
DOWEH (%) -0.024*
0.010 -2.511 0.012
Duration of Married life (DURM), (Year) 0.010 0.013 0.780 0.435
Spousal Age Gap (SAGEG) (Year) -0.022 0.027 -0.812 0.417
Husband’s Education (HEDU) (Year) -0.054*
0.033 -1.632 0.103
Highest education among male household members
HIMEDU(Year) -0.078
* 0.032 -2.408 0.016
Type of Family, TYFAMI (Nuclear = 1) -0.162
0.278 -0.582 0.561
Household Occupation CULTI, (Cultivation =1)
-0.659*
0.278 -2.372 0.018
Household Occupation NONFARM (Non-Farm= 1) -0.676**
0.347 -1.945 0.052
Household’s Land holding, HLAND, (bigha) 0.089**
0.044 2.037 0.042
Dependency Ratio in the Household DRATIO (%) 0.004 0.005 0.822 0.411
Per Capita Household Income (APCHIN) (Rs. ‘000) -0.024**
0.012 -2.063 0.039
Dowry Given at Marriage (DOW)(Yes=1) 1.434*
0.287 4.994 0.000
Post Marriage Dowry Demand (PMDOW) (Yes=1) 1.366*
0.345 3.964 0.000
Drug Addiction of the Husband (ADDIC) (Yes=1) 0.885*
0.210 4.223 0.000
Community Characteristics
DOWEC (%) -0.014 0.011 -1.368 0.171
Duration of SHG-membership DSHGM (Month) 0.007**
0.003 2.202 0.028
Caste (OBC=1) 0.098 0.269 0.364 0.716
Caste (SC=1) 0.436 0.283 1.537 0.124
Caste (ST=1) 1.064*
0.425 2.502 0.012
Summary Statistics
LR statistic (19 d f) (Probability) 175.734 (0.000) Akaike information Criterion 1.136
McFadden R-squared 0.221 Schwarz Criterion 1.286
Source: Author’s own computation based on sample observations, 2012-13 *, ** and *** imply that coefficients are significant at level 1%, 5% and 10% respectively.
The estimated coefficient of the duration of married life and the coefficient of spousal
age gap have appeared statistically insignificant. Besides, direction of the effect of the
duration of married life, which was expected as a reducing factor of domestic violence
against women, is rambling in determination of the probability of the incidence of
domestic violence. We also find that higher spousal age gap reduces the incidence of
domestic violence which is also unexpected. However, we are not worried about these
two findings because these empirical relations are statistically insignificant.
170
Table-5.3.2 Marginal Probability for the Incidence of Domestic Violence When
Women’s Empowerment is Simple Average of the Indicators
Marginal Effects After Logit
Dependent Variable (y) = Probability for the Incidence of Domestic Violence (PDVIO) (predict) =0.5922
Included observations: 580
Explanatory Variable
dy/dx
evaluated
at mean
Std.
Error
z-
Statistic Prob.>|z|
Mean of the
Explanatory
Variable(x)
Personal/Household Characteristics
DOWEH (%) -0.0058**
0.0023 -2.5200 0.0120 68.2120
Duration of Married life (DURM), (Year) 0.0024 0.0031 0.7800 0.4350 16.8690
Spousal Age Gap (SAGEG) (Year) -0.0054 0.0066 -0.8100 0.4170 5.5940
Husband’s Education (HEDU) (Year) -0.0131***
0.0080 -1.6300 0.1030 4.8670
Highest education among male household members
HIMEDU(Year) -0.0187* 0.0078 -2.4100 0.0160 6.9930
Type of Family, TYFAMI (Nuclear = 1) # -0.0387 0.0657 -0.5900 0.5560 0.8180
Household Occupation CULTI, (Cultivation =1) #
-0.1588* 0.0661 -2.4000 0.0160 0.4430
Household Occupation NONFARM (Non-Farm= 1) # -0.1667
** 0.0854 -1.9500 0.0510 0.1430
Household’s Land holding, HLAND, (bigha) 0.0214**
0.0105 2.0300 0.0420 2.6460
Dependency Ratio in the Household DRATIO (%) 0.0010 0.0012 0.8200 0.4110 49.6840
Per Capita Household Income (APCHIN) (Rs. ‘000) -0.0058**
0.0028 -2.0500 0.0400 13.7884
Dowry Given at Marriage (DOW)(Yes=1) # 0.3431
* 0.0639 5.3700 0.0000 0.6910
Post Marriage Dowry Demand (PMDOW) (Yes=1) # 0.2816
* 0.0549 5.1300 0.0000 0.1568
Drug Addiction of the Husband (ADDIC) (Yes=1) # 0.2137
* 0.0504 4.2400 0.0000 0.5200
Community Characteristics
DOWEC (%) -0.0035 0.0026 -1.3700 0.1720 54.3760
Duration of SHG-membership DSHGM (Month) 0.0017**
0.0008 2.2000 0.0280 27.2390
Caste (OBC=1) # 0.0235 0.0642 0.3700 0.7140 0.2320
Caste (SC=1) # 0.1034 0.0657 1.5700 0.1160 0.3360
Caste (ST=1) # 0.2261
* 0.0752 3.0100 0.0030 0.1170
(#) dy/dx is for discrete change of dummy variable from 0 to 1
*, ** and *** imply that coefficients are significant at level 1%, 5% and 10% respectively. Source: Author’s own computation based on sample observations, 2012-13
The coefficient of husband’s education is -0.054. It confirms that with higher educated
husband women gain strong foothold to fight against domestic violence. The estimation
of marginal probability also supports this result. The probability of facing domestic
violence reduces by 1.3% if the husband is educated one year more from mean education
level. Highest education among the male members in the family also reduces the
probability of the incidence of domestic violence against women. One year extra
schooling of the highest qualified male person in the family from mean education level
reduces the probability of having experience of domestic violence by 1.8% points.
171
The coefficient of the dummy for family type is negative but insignificant. We have
measured the impact of household occupations with reference to wage labour class. The
coefficient of major household occupations CULTI and NONFARM are negative and
statistically significant at 1.8% and 5% level of significance respectively. It implies that
the incidence of domestic violence among women belonging to cultivator family and
non-farm employment holder family is less than that among the women belonging to
wage labour class. In table-5.3.2 we see that the probability of experiencing domestic
violence among the women belonging to cultivator family is 15% lower than that among
the labour class women. If a woman belongs to non-farm employment family instead of
belonging to the wage labour family the chance of domestic violence will reduce by 16%
points. In case of cultivator family and non-farm employment family women have some
control over asset and have some decision making power. Whereas in case of labour
family women may have some earning but it is controlled by family members and
usually they have no physical assets. As a result women belonging to labour class face
more violence compared to others.
The estimate of domestic violence shows that household’s landholding is a significant
determinant of the incidence of domestic violence against women. One bigha extra
landholding over the mean landholding increases the probability of domestic violence by
2.1% points. Landholding is a strong indicator of economic status at least in rural areas.
Higher size of landholding by a family implies the higher social and economic status of
that family. However, in rural area most of the women have no ownership of land. So,
higher size of household’s landholding means higher inequality in asset holding against
women. It often makes women exclusively dependent on male persons. This factor
makes domestic violence easy against women. Estimate of the logit model shows that
dependency ratio is a stimulating factor of domestic violence. Ten percent additional
dependency ratio increases the probability of domestic violence by one percent point.
This result supports our hypothesis but it is not significant.
We have found that higher per capita family income reduces the probability of
experiencing domestic violence against women, this relation is statistically significant. In
table-5.3.2 we see that one thousand additional household’s annual per capita income
above mean reduces the probability of experiencing domestic violence by 0.5% point.
Higher household income means higher economic status and occupational status and
172
higher access to asset and wealth. These facts usually give honour the women in the
family as we expect. Our empirical analysis strongly supports our expectation.
Therefore, poor women in terms of land and income are more victimised in domestic
violence compared to others.
We have found that a major section of sample women has given dowry at marriage. The
coefficient of dowry given at marriage tells that women whose houses paid dowry at the
time marriage face more risk of domestic violence. The probability of facing domestic
violence for the women who paid dowry is 34% greater than the probability of facing
domestic violence for the women who did not pay dowry. Further, women who are
forced to pay extra dowry after marriage are more likely to face domestic violence. This
result is statistically significant at 1% level. In table-5.3.2 we find that post marriage
dowry demand increases the probability of domestic violence by 28%. Usually, women
are not willing to bring dowry from her natal house after marriage. This disagreement
often creates quarrel between the woman and in-laws or husband and thereby the women
face domestic violence. So, the result is justified. We have got that the coefficient of
drug addiction of the husband is positive and statistically significant. Drug addition of
the husband increases the probability of facing domestic violence 21% points. Usually
women protest against drug addiction of the husband. But addiction does not tolerate any
protest and thereby make violence against women. Our findings agree with this view.
Therefore, dowry at the marriage and post marriage time and drug addiction of the
husband are the main cause of domestic violence in the area under study.
We have observed that the community level empowerment of women reduces the
probability of domestic violence. The marginal change of probability reports that one
percent higher degree of community level empowerment above mean reduces the
probability of facing domestic violence by 0.35% but the result is statistically significant
at 17% level. It establishes that household level empowerment is more important that the
community level empowerment of the women for combating the curse of domestic
violence against them.
The coefficient of the duration of SHG membership is positive and statistically
significant at 2% level. Our empirical result shows that probability of domestic violence
will increase by 2% points if the duration of SHG membership increases by one year
173
from average. Therefore, duration of SHG membership has been found to stimulate the
probability of suffering from domestic violence. The marginal probability of domestic
violence against women increases due to increase of the duration of SHG membership in
the area under study. This result goes against our hypothesis. This empirical result tells
us that women participate in SHG movement at the cost of domestic violence. By local
customs households don’t want that their women will move outside home and form a
group for social and economic movement. Sometimes women become member of SHGs
even when their family members oppose to do it. As a result there occurs domestic
violence within the family. So, domestic violence against women is an impediment of
financial inclusion for women in rural Bankura.
We have considered the women belonging to general caste as base category for assessing
the impact of caste dummies on the probability of facing domestic violence. The
coefficient of caste (OBC =1) is positive but insignificant. The coefficient of Caste
(SC=1) is positive and statistically significant at 11% level. This result is indicative that
the women belonging to scheduled castes are suffering more from domestic violence in
contrast to the women of general castes. Moreover, the coefficient of the caste dummy
(ST=1) is positive and statistically significant. The marginal probability calculation
shows that the probability of experiencing domestic violence is 22% (10%) higher for
the Scheduled tribe (scheduled castes) women compared to general caste women. During
our field survey we have noticed that there exist several types of superstitions among the
people of ST community such as Daini (witch) etc. which provokes violence against
women within the family. Therefore, scheduled tribes women are more vulnerable in
terms of domestic violence in the area under study.
We, therefore, conclude that women’s empowerment at the household level is an
important determinant of the probability of facing domestic violence. Moreover, we have
found that husband education, household occupation, dowry at marriage and at post
marriage and husband’s drug addiction are effective factors of domestic violence. This
empirical estimation, however, reveals that community level empowerment of women,
duration of married life, spousal age gap, type of family, dependency ratio and caste SC
and Caste OBC are statistically insignificant in the determinant of the probability of
domestic violence against women in the area under study.
174
5.3.2. Model-2B: Logit Model with Composite Empowerment Indices
In this model composite index of women’s empowerment at the household level and at
the community level have been considered as exogenous variables along with the other
exogenous variables (as they were in Model-2A) for estimating the impact of
empowerment on the incidence of domestic violence. Table-5.3.3 and table-5.3.4
represent the results of the Model-2B.
Table-5.3.3 Results of the Logit Model for the Incidence of Domestic Violence
When Women’s Empowerment is Composite Index of the Indicators
Dependent Variable: Incidence of Domestic Violence (DVIO)
Method: ML - Binary Logit (Newton-Raphson) Included observations: 580
Convergence achieved after 4 iterations
Covariance matrix computed using second derivatives
Explanatory Variable Coefficient Std.
Error z-Statistic Prob.
Constant -0.736 0.727 -1.013 0.311
Individual/household Characteristics
CIWEH -0.018*
0.006 -2.781 0.005
Duration of Married life (DURM), (Year) 0.010 0.013 0.830 0.407
Spousal Age Gap (SAGEG) (Year) -0.022 0.027 -0.827 0.408
Husband’s Education (HEDU) (Year) -0.056***
0.033 -1.693 0.091
Highest education among male household members
HIMEDU(Year) -0.077
* 0.032 -2.397 0.017
Type of Family, TYFAMI (Nuclear = 1) -0.234 0.278 -0.845 0.398
Household Occupation CULTI, (Cultivation =1)
-0.641*
0.277 -2.313 0.021
Household Occupation NONFARM (Non-Farm= 1) -0.659**
0.346 -1.903 0.057
Household’s Land holding, HLAND, (bigha) 0.088**
0.043 2.032 0.042
Dependency Ratio in the Household DRATIO (%) 0.005 0.005 0.911 0.362
Per Capita Household Income (APCHIN) (Rs. ‘000) -0.024*
0.011 -2.087 0.037
Dowry Given at Marriage (DOW)(Yes=1) 1.403* 0.287 4.893 0.000
Post Marriage Dowry Demand (PMDOW) (Yes=1) 1.363*
0.343 3.979 0.000
Drug Addiction of the Husband (ADDIC) (Yes=1) 0.902*
0.210 4.303 0.000
Community Characteristics
CIWEC -0.007 0.007 -0.975 0.330
Duration of SHG-membership DSHGM (Month) 0.006**
0.003 2.005 0.045
Caste (OBC=1) 0.084 0.268 0.312 0.755
Caste (SC=1) 0.428 0.283 1.516 0.129
Caste (ST=1) 1.053*
0.426 2.474 0.013
Summary Statistics
LR statistic (19 d f) (Probability) 172.282 (0.000) Akaike information Criterion 1.142
McFadden R-squared 0.217 Schwarz Criterion 1.292
Source: Author’s own computation based on sample observations, 2012-13
*, ** and *** imply that coefficients are significant at level 1%, 5% and 10% respectively.
175
Table-5.3.4 Marginal Probability of the Incidence of Domestic Violence When
Women’s Empowerment is Composite Index of the Indicators
Marginal Effects After Logit
Dependent Variable (y) = Probability for the Incidence of Domestic Violence (PDVIO) (predict)
= 0.59081508
Included observations: 580
Explanatory Variable
dy/dx
evaluated
at mean
Std.
Error
z-
Statistic Prob.>|z|
Mean of the
Explanatory
Variable(x)
Individual/Household Characteristics
CIWEH -0.004* 0.0016 -2.7800 0.0050 3.00E-06
Duration of Married life (DURM), (Year) 0.002 0.0030 0.8300 0.4070 16.8690
Spousal Age Gap (SAGEG) (Year) -0.005 0.0066 -0.8300 0.4080 5.5948
Husband’s Education (HEDU) (Year) -0.0135***
0.0080 -1.6900 0.0900 4.8672
Highest education among male household members
HIMEDU(Year) -0.0185
* 0.0077 -2.4000 0.0170 6.9931
Type of Family, TYFAMI (Nuclear = 1)# -0.0557 0.0648 -0.8600 0.3890 0.8190
Household Occupation CULTI, (Cultivation =1) #
-0.1546* 0.0661 -2.3400 0.0190 0.4431
Household Occupation NONFARM (Non-Farm= 1) # -0.1626
** 0.0852 -1.9100 0.0560 0.1431
Household’s Land holding, HLAND, (bigha) 0.0213**
0.0105 2.0300 0.0420 2.6465
Dependency Ratio in the Household DRATIO (%) 0.0011 0.0012 0.9100 0.3620 49.6845
Per Capita Household Income (APCHIN) (Rs. ‘000) -0.0059**
0.0028 -2.0900 0.0370 13.7884
Dowry Given at Marriage (DOW)(Yes=1) # 0.3363
* 0.0642 5.2400 0.0000 0.6914
Post Marriage Dowry Demand (PMDOW) (Yes=1) # 0.2817
* 0.0549 5.1300 0.0000 0.1569
Drug Addiction of the Husband (ADDIC) (Yes=1) # 0.2181
* 0.0507 4.3000 0.0000 0.5207
Community Characteristics
CIWEC -0.0016 0.0017 -0.9700 0.3300 1.10E-06
Duration of SHG-membership DSHGM (Month) 0.0015**
0.0008 2.0100 0.0450 27.2397
Caste (OBC=1) # 0.0202 0.0643 0.3100 0.7540 0.2328
Caste (SC=1) # 0.1018 0.0656 1.5500 0.1210 0.3362
Caste (ST=1) # 0.2247
* 0.0758 2.9600 0.0030 0.1172
(#) dy/dx is for discrete change of dummy variable from 0 to 1 *, ** and *** imply that coefficients are significant at level 1%, 5% and 10% respectively.
Source: Author’s own computation based on sample observations, 2012-13
In the model of the incidence of domestic violence where empowerment is measured by
the composite index of the indicators, we have found almost the same results as we have
found in Model-2A (table-5.3.1 and table-5.3.2). Table-5.3.4 shows that one percent
increase in women’s empowerment at the household level from its mean reduces the
probability of the incidence of domestic violence by 0.43% point. However, the
coefficient of community level empowerment of woman is statistically insignificant.
Therefore, in this model we also observed that household level empowerment is effective
176
to reduce domestic violence against women whereas community level empowerment is
not so important for reducing domestic violence. In addition to the household level
empowerment of women we find that husband education and education of other family
member reduce the incidence of domestic violence for the sample women. On the other
hand, dowry demand, drug addiction and participation in SHG stimulate domestic
violence against women in the district of Bankura, West Bengal. Therefore, change of
methodology for estimating the impact of women’s empowerment did not alter our
findings regarding the incidence of domestic violence.
5.4. Impact of Women’s empowerment on Expenditure for Child Education
In this section, we have interpreted the estimated results of child education expenditure
as proportion to annual household income. It is quite natural that in our sample all
women do not have children of school going age. So the estimation of the impact of
women’s empowerment on child education expenditure as proportion to household
income would be on the basis of those sample members who have school aged children
during the time of interview. Although our study is based on 580 observations, only 431
sample women have school aged children, who are either enrolled or not in school.
Therefore, the estimation of child education expenditure as proportion to annual
household income is based on only 431 observations. The issue of child education
expenditure as proportion to annual household income has been estimated using two log-
lin models, Model-3A and Model-3B. In Model-3A we have included the degrees of
women’s empowerment at the household level and at the community level in addition to
some selected households and community characteristics as exogenous variables. We
have considered the composite index of women’s empowerments including the other
exogenous variables in the determination of child education expenditure as proportion to
annual household income in Model-3B. Among other variables, we have incorporated
education level of father and other household members, household’s land holding,
household occupation, dependency ratio, composition of family, annual per capita family
income in the range of individual and household characteristics. Caste and the duration
of SHG-membership have been included as other community characteristics. The results
of the Model-3A and Model-3B have been presented in table-5.4.1 and in table-5.4.2
respectively. In these log-linear models we have a few dummy independent variables.
We have interpreted the coefficients of the dummies following the formula of Halvorsen
and Palmquist, (1980) [{exp (coefficient of dummy variable) – 1} x 100]. The summary
177
statistics shows that the log-linear models for child education expenditure are good fitted
and there is no heteroscedasticity problem in the estimates.
In table-5.4.1 the coefficient of women’s empowerment at the household level is 0.0014.
This result supports our expected direction but it is statistically insignificant. If we look
into the coefficient of community level empowerment, we find that community level
empowerment has a positive and significant impact on the proportion of household
income on child education. It implies that one percent increase in the degree of women’s
empowerment at the community level increases the proportion of household income
spent on child education by 0.48%. For Model-3B, this line of findings has also been
confirmed. It means that although both the household and community level
empowerment are instrumental for spending on child education, the community level
empowerment of women is more fruitful for spending more on child education.
Empowerment at the community level increases the spending in three ways. First, a
woman always wants that her child would be educated. This want is more active for
empowered women. So, it is expected that empowered women at the community level
spend or force to spend more of household income for her child. Second, empowered
women have own decision making power regarding the matters relating to children. This
power can increase expenditure for her children. Third, empowered women at the
community level are more conscious of the education of their children. This
consciousness induces to increase the expenditure for children out of school like private
tuition, school uniform, expenditure for co-curricular activities etc.
From the sign of the coefficient of father’s education (husband’s education, HEDU) we
can say that educated father spend smaller percentage of household income for his child
education. This result is statistically significant at one percent level. In table-5.4.1 and in
table-5.4.2 we find that one year extra schooling of father reduces the share of household
income spending for child education by 2.8%. Apparently, one may argue that this result
is unexpected. But this result may be supported by some economic logic. It is expected
that educated father earns more; so it is expected that as income increases due to higher
education of father the share of education expenditure will reduce. Therefore, our
empirical finding regarding the impact of father education on spending for child
education has a logical base.
178
Table-5.4.1 Results of the Log-Linear Model for Child Education When Women’s
Empowerment is Simple Average of the Indicators
Dependent Variable: LEDEX{ln(Child Education Expenditure as Proportion to Annual Household
Income)}
Method: Least Squares
Included observations: 431
White heteroskedasticity-consistent standard errors & covariance
Explanatory Variable Coefficient Std. Error t-Statistic Prob.
Constant 1.8434 0.4438 4.1537 0.0000
Individual/household Characteristics
DOWEH (%) 0.0014 0.0017 0.8179 0.4139
Father’s education (HEDU) (Year) -0.0285*
0.0059 -4.8157 0.0000
Highest education among male household members,
HIMEDU,(Year) 0.0361* 0.0063 5.7362 0.0000
Highest education among female household
members HIFEDU(Year) 0.0217* 0.0056 3.8770 0.0001
Type of Family TYFAMI, (Nuclear =1) 0.2271* 0.0481 4.7262 0.0000
Household Occupation CULTI, (Cultivation =1)
0.0031 0.0471 0.0649 0.9483
Household Occupation NONFARM (Non-Farm= 1) 0.0371 0.0623 0.5951 0.5521
Household’s Land holding, HLAND, (bigha) 0.0058 0.0073 0.7896 0.4302
Dependency Ratio in the Household DRATIO (%) 0.0048* 0.0012 4.0110 0.0001
ln(APCHIN) -0.2408* 0.0493 -4.8827 0.0000
Community Characteristics
DOWEC (%) 0.0048* 0.0018 2.6326 0.0088
Duration of SHG-membership DSHGM (Month) 0.0004 0.0005 0.7179 0.4732
Caste (OBC=1) -0.0437 0.0461 -0.9484 0.3435
Caste (SC=1) -0.0880***
0.0521 -1.6891 0.0920
Caste (ST=1) -0.0938 0.0729 -1.2861 0.1991
Summary Statistics
R-squared 0.3402 Akaike information criterion 0.7690
Adjusted R-squared 0.3147 Schwarz criterion 0.9294
Durbin-Watson statistic 0.7178 F-statistic (Probability) 13.339 (0.000) Source: Author’s own computation based on sample observations, 2012-13
*, ** and *** imply that coefficients are significant at level 1%, 5% and 10% respectively.
The coefficient of highest education among male household members in family is
positive and statistically significant. The magnitude of the coefficient speaks that one
year extra education of the highest qualified male member in the households increases
the proportion of household income spent on child education by 3.6% point. We have
got the same effect of highest education among male household members for Model-3B.
This result supports our hypothesis. The qualified male and female members understand
better the importance of education and accordingly oblige the parents to spend more for
the children. We have also found that highest education among female household
members have also positive and significant impact on child education expenditure as
proportion to household income. One year extra education of the highest qualified
179
female increases share of child education expenditure in income by 2.1% point for both,
Model-3A and in Model-3B. We have observed that effect of male education is greater
than the effect of female education on the share of child education expenditure in
household income. It may happen due to the fact that the average education of the
highest qualified male is greater than the education of the highest qualified female. In the
course of field survey we have generally observed that highest educational qualification
goes in favour of the younger members who are still studying or seeking jobs. Naturally,
parents are inspired for spending more out of their income for their children’s education.
Therefore, educational back ground of the household is an important determinant of the
share of child education expenditure in household income.
Table-5.4.2 Results of the Log-Linear Model for Child Education When Women’s
Empowerment is Composite Index of the Indicators
Dependent Variable: LEDEX{ln(Child Education Expenditure as Proportion to Annual Household
Income)}
Method: Least Squares
Included observations: 431
White heteroskedasticity-consistent standard errors & covariance
Explanatory Variable Coefficient Std. Error t-Statistic Prob.
Constant 2.2103*
0.4314 5.1235 0.0000
Individual/Household Characteristics
CIWEH 0.0016 0.0012 1.3933 0.1643
Father’s education (HEDU) (Year) -0.0282* 0.0060 -4.7438 0.0000
Highest education among male household members,
HIMEDU,(Year) 0.0361* 0.0063 5.6883 0.0000
Highest education among female household
members HIFEDU(Year) 0.0218* 0.0056 3.9008 0.0001
Type of Family TYFAMI, (Nuclear =1) 0.2343* 0.0479 4.8936 0.0000
Household Occupation CULTI, (Cultivation =1)
0.0071 0.0472 0.1506 0.8804
Household Occupation NONFARM (Non-Farm= 1) 0.0406 0.0625 0.6484 0.5171
Household’s Land holding, HLAND, (bigha) 0.0061 0.0073 0.8373 0.4029
Dependency Ratio in the Household DRATIO (%) 0.0047* 0.0012 3.8351 0.0001
ln(APCHIN) -0.2430* 0.0489 -4.9680 0.0000
Community Characteristics
CIWEC 0.0027* 0.0011 2.3576 0.0189
Duration of SHG-membership DSHGM (Month) 0.0005 0.0005 0.9632 0.3360
Caste (OBC=1) -0.0397 0.0464 -0.8554 0.3928
Caste (SC=1) -0.0815 0.0528 -1.5422 0.1238
Caste (ST=1) -0.0866 0.0725 -1.1955 0.2326
Summary Statistics
R-squared 0.3376 Akaike information criterion 0.7729
Adjusted R-squared 0.3120 Schwarz criterion 0.9333
Durbin-Watson statistic 0.7115 F-statistic (Probability) 13.187 (0.000) Source: Author’s own computation based on sample observations, 2012-13
*, ** and *** imply that coefficients are significant at level 1%, 5% and 10% respectively.
180
According to our hypothesis the coefficient of the dummy for type of family (1= Nuclear
family) in Model-3A and in Model-3B are expected to be positive. In the empirical
estimation these coefficients are 0.227 in the Model-3A and 0.234 in Model-3B. These
results are statistically significant at 1% level. These indicate that the child education
expenditure as proportion to household income for the nuclear families is greater than
that for the joint or extended families. Following the formula of Halvorsen and
Palmquist, (1980) the coefficient of the dummy for type of family in the Model-3A
indicates that the mean child education expenditure as proportion to the household
income is 25.49% higher for the nuclear family than that for the other types of family.
Qualitatively almost same results we have found for Model-3B. These results support
our hypothesis. Therefore, we come to the conclusion that the nuclear composition of
family increases the child education expenditure as proportion to household income in
the district of Bankura. During the field survey we have seen that most of the parents in
nuclear family compared to joint and extend families are more serious and have soul
authority to spend for their child. It justifies our result regarding the effect of family
composition on child education expenditure.
In order to assess the impact of household occupation on pattern of spending on child
education we have categorized the household occupation into three categories –
cultivation, non-farm self employment/service and wage labour class. For econometric
analysis labour class has been considered as reference category. Table-5.4.1 and table-
5.4.2 exhibit that the household occupations (Cultivation=1) and (Non-farm self
employment=1) have positive impact on child education expenditure. We find that in
both the models child education expenditure as proportion to household income is higher
for cultivator family (0.31%) and for self-employed/service holder family (3.77%) than
labour class family. However, the impacts of household occupation on child education
expenditure are not statistically significant. The coefficient of landholding in both the
models indicates that landholding has a direct effect on child education expenditure. It is
compatible with our expectation but this result is statistically insignificant.
This study asserts that dependency ratio has some favorable effect on child education
expenditure as proportion to household income. The estimated coefficient shows that one
percent increase in dependency ratio increases the share of child education expenditure
by 0.48% in Model-3A (0.47% in Model-3B). This result is statistically significant at 1%
181
level. Although this finding goes against our hypothesis we can explain the findings
based on our ground experiences. In the course of field survey we have observed that a
major portion of the sample households are nuclear and dependents are children.
Therefore, the higher is the value of dependency ratio the higher is the number of
children. Number of children definitely increases the share of income spent on child
education. That is why, we have obtained a direct relation between dependency ratio and
child education expenditure as proportion to household income.
We know that any kind of expenditure depends on income. Like other expenditure child
education expenditure necessarily depends on household income. In our log-linear model
we have tried to examine the effect of the log of per capita income on the log of child
education expenditure as proportion to household income. For both Model-3A and
Model-3B, we find that the coefficient of the log of per capita household income is –
0.24 which is statistically significant at one per cent level. It indicates that one per cent
increase in per capita household income reduces the share of income spent one child
education by 0.24%.
Now we interpret the impact of community characteristics on child education
expenditure. We have already explained that community level empowerment of women
have a significant impact on child education expenditure as proportion to household
income. We have observed that a large section of household particularly women have
joined SHG-centric microfinance which organize the poor women for financial inclusion
and let them to understand the importance of child education and health and other social
issues. With this end in view, we have taken the duration of SHG membership as an
explanatory variable in the spectrum of community characteristics in this model of child
education expenditure. The coefficient of the duration of SHG-membership is found to
be positive. It means that if any member in household participates in SHG and continues
the membership, child education expenditure increases. It is as per with our expectation.
However, this empirical result in both model are statistically insignificant. It is indicative
that SHG movement may be successful to ensure financial inclusion but it is less
important to enhance the expenditure for child education.
Like other issues regarding women’s empowerment in the estimation of child education
expenditure as proportion to household income we have considered three dummies for
182
castes. The households belonging to general castes are reference category. The
coefficients of the caste dummies are negative. They indicate that the share of household
income spent for child education is lower for the OBC, SC and ST households in
contrast to the general caste households. The coefficient of the dummy for caste (SC=1)
is statistically significant at 9% level in Model-3A and at 12% level in Model-3B. We
compute that child education expenditure as proportion to household income for
scheduled castes household is 8.42 % smaller than that for general castes households.
However, the coefficient of the dummies Caste (OBC=1), Caste (ST=1) are not
statistically significant. The backwardness of the lower castes and unconsciousness
regarding child education gives these results. It makes the vicious circle of educational
poverty. Due to backwardness and income poverty, the lower castes could not spend
more for education which in turn keep these people backward in terms of education and
income.
This section, therefore, concludes that women’s empowerment i.e. mothers
empowerment, father’s education, highest education of the male persons in the family,
highest education of the female persons in family, family type, dependency ratio, income
and caste are the most important determinants of child education expenditure as
proportion to household income for the households in the district of Bankura.
5.5. Determinants of Women’s Empowerment in Bankura District
In the last three subsequent sections, we have interpreted and explained the impact of
women’s empowerment at the household level and at the community level along with
selected household and community characteristics on three issues of household and child
welfare. We have found that women’s empowerment at the household level and at the
community level have some positive and significant effect on the probability of adopting
family planning decision for the households in the district of Bankura. Women’s
empowerment at the household level and at the community level reduces the probability
of the incidence of domestic violence against women. Our empirical research has also
shown that women’s empowerment is very much important on spending more for child
education. Therefore, it has been empirically established that women’s empowerment are
instrumental for household and child welfare in the district of Bankura. Once we have
found the instrumental role of women’s empowerment, we should examine the
determining factors of women’s empowerment at the household level and at the
183
community level. To this end, we have estimated two multiple linear regression models –
one for women’s empowerment at the household level and another for women’s
empowerment at the community level. In this section, with two sub-sections we interpret
the estimated models relating to women’s empowerment at the household level and at
the community level.
5.5.1. Determinants of Women’s Empowerment at the Household Level
In this sub-section, we explain the estimated multiple linear regression model relating to
women’s empowerment at the household level. Based on two measures of women’s
empowerment at the household level, we have estimated two multiple liner regression
models with same set of explanatory variables. Both these multiple regression models
include some quantitative explanatory variables and some dummy variables. This
implies that our regression models relating to the women’s empowerment are known as
Analysis of Covariance Models (ANCOVA). First, we discuss the findings of the model
(Model-4A) where household level women’s empowerment has been measured by
simple average of its Indicators. Next we explain the models (Model-4B) of household
level women’s empowerment measured by weighted average of PCA of the indicators.
The empirical estimates of models have been presented in table-5.5.1 and table-5.5.2
respectively. The goodness of fit is reasonable for both the models. Again F-statistics is
significant for both the model. The result of White test of heteroscedasticity reveals that
estimates of the models for household empowerment of women do not suffer from
heteroscedasticity problem.
Age is a very important factor for growing personality of a person. Although age of a
person is quantitative variable, in this study sample women have been divided into four
age groups for examining the impact of age on their empowerment. We have included
three dummies and the older age group (age>45 Years) has been considered as reference
category. The coefficient of age group (< 25 years) is positive. It indicates that
controlling other explanatory variables, the age group (< 25 years) enjoys 3% point
higher empowerment at the household level compared to the household level
empowerment of the older group. But the result is statistically significant at 14.5 % level.
The age group AGE2 (25-35 years) of the women has positive and significant impact on
empowerment index. The average degree of empowerment of women at the household
level increases by 4.26% points if the women belong to AGE2 group instead of AGE4.
184
Further, women under age group AGE3 (36-45 Years) enjoy 3% points higher
empowerment at the household level compared to the empowerment at the household
level of the reference age group. In Model-4B we have got the qualitatively same result
regarding the effect of age on women’s empowerment at the household level. We have,
therefore, found that young women in contrast to the older women are more empowered.
However, it does not mean the reduction of traditional respects of elder rather now a-
days elders share their power with the younger.
The coefficient of the education level of the women is negative in Model-4A and
positive in Model-4B. The result of Model-4B justifies our hypothesis. But both the
results are statistically insignificant. It is indicative that women’s education is not so
much important in the determination of the women’s empowerment at the household
level in the district of Bankura. Education of women should have positive and significant
impact on household level women’s empowerment. Yet we should point out the fact that
most of the sample women have education below primary level. So, what impact should
we expect from this level of education on empowerment? Further, in the time of
interview we have observed that many educated women could not take several familial
decisions and don’t have any say regarding economic matter of the households. On the
other hand, many illiterate women in the low caste family enjoy commendable
empowerment at the household level.
Personal occupation of the women is, no-doubt, a crucial determinant of women’s
empowerment at the household level. Personal occupation of our sample women has
been categorized in three groups – wage labour, self-employed or formal service holders
and homemakers. Homemaker is here benchmark category. Our empirical estimates
shows that the coefficient LAB (Wage labour =1) is negative and statistically significant.
Controlling the other things, the magnitude of the coefficient of LAB (Wage labour =1)
signifies that average household level empowerment of wage labour women is 6.1%
point lower than that of the homemakers. It goes against our hypothesis. Wage labourers
earn some income but in most of the cases earnings are controlled by their families,
particularly, by their husbands. Besides, wage earners are usually illiterate or just
literates, accordingly they are less concern regarding the economic decisions. It may
explain the causes of the low level household empowerment of the wage labourers
compared to homemakers. On the other hand, the coefficient of SELF (Self employed or
185
service=1) is positive (4.26) and statistically significant. It says that mean of household
level empowerment of self employed or service holder women is higher by about 4.26%
point than mean of household level empowerment of the benchmark category. This result
has justified our hypothesis. Relative to common homemaker self employed and formal
service holders are definitely more powerful to take several household level decisions
and hold some more physical and financial asset. In Model-4B, we have obtained almost
same findings regarding the impact of personal occupation on household level
empowerment of the women.
Estimating the ANCOVA model for women’s empowerment at the household level we
find that personal income of the women has some positive and statistically significant
impact on household level empowerment of women. The coefficient of personal income
shows that one thousand additional average annual personal income of the woman
increases the household level empowerment by 9.36% points from average. In Model-4B
we have seen the qualitatively same result regarding the impact of personal income on
household level empowerment. Therefore, our empirical results support the hypothesis.
When a woman earn, she becomes economically independent. We have projected that
income of wage labor women is controlled by other family members. But in average for
all we can say that women earners have some economic dignity or empowerment in their
households.
In order to assess the impact of financial inclusion on household level empowerment of
the women we have included the dummy AFCT (access to formal credit, Yes=1) as an
explanatory variable in the model for empowerment at the household level. Our
empirical estimates of the ANCOVA models, Model-4A and Model-4B, show that the
coefficient of AFCT is positive and statistically significant. The coefficient refers to that
other factors remaining same if a woman has access to formal credit it increases her
empowerment at the household by 4.45% points from the average. This result supports
our hypothesis and is not hard to explain the logic behind. Access to formal credit
enhances the importance of the women in her household through borrowing from
institution in different difficult situations. Further, when a woman can borrow from
formal institution it is expected that she is powerful in different familial decision making
process. Having access to formal credit, women may take decisions regarding children’s
education and saving credit decision independently. So, it is expected that access to
186
formal credit by women inculcates women’s empowerment. Therefore, financial
inclusion increases the household and child welfare through the channel of women’s
empowerment at the household level.
We obtain that the coefficient of the dummy for the type of family is 4.99 in Model-4A
and it is 5.45 in Model-4B. Therefore, the coefficient of the dummy for family type tells
us that household level empowerment of women of nuclear families is almost 5% points
higher from average household level empowerment of the women of joint/extended
families. It is not surprising that women in the nuclear families enjoy exclusive power in
economic and familial decision making process. So, our empirical result regarding the
effect of family type on household level empowerment of women is reasonably
conclusive.
We have observed that dependency ratio has some adverse effect on the household level
empowerment of women. In both the regression analysis the coefficient of the
dependency ratio is negative and statistically significant. In Model-4A the coefficient of
the dependency ratio interpret that one percent increase in dependency ratio from
average level reduces the household level empowerment of the women by 0.5% points
from average empowerment. However, dependency ratio reduces empowerment but the
effect is marginal. In the family with higher dependent members like old or child,
women have to spend more time in the household’s jobs. They get less time for earning
as well for knowing and thinking about various economic issues. So, higher dependency
ratio reduces women’s empowerment within the household.
In our empirical estimation we have got a favourable impact of per capita household
income on women’s empowerment at the household level in both the models– Model-4A
and Model-4B. The coefficient of the per capita household income tells us that
household level empowerment of women increases with household income. However,
the empirical relation between household income and women’s empowerment is not
statistically significant. Therefore, our study suggests that although personal income is
significantly important, household income is immaterial in the determination of women’s
empowerment at the household level for the women in the district of Bankura of West
Bengal.
187
Table-5.5.1 Estimates of degree of Women’s Empowerment at Household Level
Dependent Variable: Degree of Women’s Empowerment at Household Level (DOWEH)
Method: Least Squares
Sample: 1 580
Included observations: 580
Explanatory Variable Coefficient Std. Error t-Statistic Prob.
Constant 59.001*
2.391 24.677 0.000
Individual/Household Characteristics
Age Group (<25 Years) AGE1 (Yes=1) 2.941 2.015 1.459 0.145
Age Group (25-35 Years) AGE2 (Yes=1) 4.261* 1.512 2.819 0.005
Age Group (36-45 Years) AGE3 (Yes=1) 3.037**
1.413 2.150 0.032
Education of the Woman, EDU (Year) -0.036 0.193 -0.184 0.854
Occupation of the Woman, LAB (Wage Labour=1) -6.102* 1.132 -5.389 0.000
Occupation of the Woman, SELF (Self employed or
Service =1) 4.268
* 1.874 2.277 0.023
Average Monthly Personal Income PINC (Rs. '00) 0.078* 0.027 2.865 0.004
Access to Formal Credit AFCT (Yes=1) 4.458* 1.166 3.823 0.000
Type of Family TYFAMI (Nuclear =1) 4.998* 1.279 3.907 0.000
Dependency Ratio in the Household DRATIO (%) -0.058**
0.025 -2.313 0.021
Per Capita Household Income (APCHIN) (Rs. ‘000) 0.051 0.045 1.151 0.250
Household Occupation CULTI, (Cultivation =1)
-5.514* 1.297 -4.251 0.000
Household Occupation NONFARM (Non-Farm= 1) 1.336 1.660 0.805 0.421
Household’s Land holding, HLAND, (bigha) 0.258 0.199 1.295 0.196
Highest education among male household members,
HIMEDU,(Year) -0.173 0.139 -1.240 0.216
Highest education among female household members
HIFEDU(Year) 0.547
* 0.161 3.399 0.001
Community Level Characteristics
Duration of SHG membership DSHGM (Month) 0.021 0.017 1.291 0.197
Caste (OBC =1) 1.578 1.309 1.205 0.229
Caste (SC=1) 1.134 1.366 0.830 0.407
Caste (ST=1) -0.412 1.824 -0.226 0.821
Summary Statistics
R-squared 0.341 Akaike information criterion 7.712
Adjusted R-squared 0.316 Schwarz criterion 7.878
Durbin-Watson statistic 1.691 F-statistic (Probability) 13.744 (0.000)
White Heteroskedasticity Test: H0: Variance of random disturbance term is constant
F-statistic (Probability) 0.985(0.548) Obs×R-squared (Probability) 228.262(0.520)
Source: Author’s own computation based on sample observations2012-13
*, ** and *** imply that coefficients are significant at level 1%, 5% and 10% respectively.
We have taken two dummy variables for major household occupation in our ANCOVA
model. The coefficient of the major household occupation dummy (Cultivation=1) is –
5.51. It implies that the average degree of household level empowerment of the women
188
belonging to cultivator family is 5.51% points lower than that of the women belonging to
wage labour family. This result is statistically significant in both the models. This
empirical result does not support our hypothesis. But, this empirical finding is insightful.
During the course of data collection we have seen that women in the wage labour family
are more vocal than the women in cultivator family. Wage labour families are usually
tribal or low castes where by custom women enjoy empowerment at the household level.
Further, in the wage labour family women are earner which often inculcates
empowerment at the household level. On the other hand, the coefficient of the dummy
indicating (self employment=1) is positive in both the models. It means that the degree
of empowerment at the household level of the women under self employed family is
higher than that of the women under wage labour. However, this result is not statistically
significant at all.
The coefficient of landholding shows that other things remaining unchanged, one bigha
extra landholding increases the household level empowerment by 0.25% points in
Model-4A and 0.52% points in Model-4B. The result in Model-4B is statistically
significant at 6% level. Therefore, the estimation of the Model-4B establishes that
household’s landholding increases the household level empowerment of the women in
the district of Bankura. Large land size of the household ensures livelihood security of
the family and the women. Often women have the land ownership in the household with
higher landholdings. These matters may enhance dignity of the women and accordingly
the empowerment at the household level.
The coefficient of highest education among the male household members is statistically
insignificant as an explanatory variable in the models for women’s empowerment at the
household level. On the other hand, the coefficient of the highest education among the
female members is positive and statistically significant at 1% level. The coefficient
implies that one extra completed year of education of the highest qualified female
member increases the average degree of empowerment at the household level by 0.55%
points. We have got almost the same result in Model-4B regarding the highest female
education. Therefore, female education is more imperative than the male education in the
family for inculcating women’s empowerment at the household level. It is expected that
most of the educated women persuades the decisions making power and the personality
of the women. Hence, our result is expected.
189
Table-5.5.2 Estimates of Composite Women’s Empowerment Index at Household
Level
Dependent Variable: Composite Index of Women’s Empowerment at Household Level (CIWEH)
Method: Least Squares
Sample: 1 580
Included observations: 580
Variable Coefficient Std.
Error
t-
Statistic Prob.
Constant -11.627 3.350 -3.471 0.001
Individual/Household Characteristics
Age Group (<25 Years) AGE1 (Yes=1) 3.757 2.823 1.331 0.184
Age Group (25-35 Years) AGE2 (Yes=1) 6.190* 2.118 2.923 0.004
Age Group (36-45Years) AGE3 (Yes=1) 4.207**
1.979 2.126 0.034
Education of the Woman, EDU (Year) 0.180 0.270 0.664 0.507
Occupation of the Woman, LAB (Wage Labour=1) -7.964* 1.586 -5.020 0.000
Occupation of the Woman, SELF (Self employed/Service =1) 7.563* 2.626 2.880 0.004
Average Monthly Personal Income PINC (Rs. '00) 0.127* 0.038 3.342 0.001
Access to Formal Credit AFCT (Yes=1) 5.017* 1.634 3.071 0.002
Type of Family TYFAMI (Nuclear =1) 5.453* 1.792 3.043 0.003
Dependency Ratio in the Household DRATIO (%) -0.089* 0.035 -2.533 0.012
Per Capita Household Income (APCHIN) (Rs. ‘000) 0.061 0.063 0.969 0.333
Household Occupation CULTI, (Cultivation =1)
-8.544* 1.817 -4.702 0.000
Household Occupation NONFARM (Non-Farm= 1) 2.666 2.325 1.147 0.252
Household’s Land holding, HLAND, (bigha) 0.521**
0.279 1.866 0.063
Highest education among male household members,
HIMEDU,(Year) -0.207 0.195 -1.059 0.290
Highest education among female household members
HIFEDU(Year) 0.488
** 0.225 2.165 0.031
Community Level Characteristics
Duration of SHG membership DSHGM (Month) 0.052**
0.023 2.237 0.026
Caste (OBC =1) 2.979***
1.834 1.624 0.105
Caste (SC=1) 2.435 1.914 1.272 0.204
Caste (ST=1) -0.085 2.555 -0.033 0.974
Summary Statistics
R-squared 0.345 Akaike information criterion 8.387
Adjusted R-squared 0.321 Schwarz criterion 8.552
Durbin-Watson statistic 1.662 F-statistic (Probability) 14.019 (0.000)
White Heteroskedasticity Test: H0: Variance of random disturbance term is constant
F-statistic (Probability) 1.224 (0.045) Obs×R-squared (Probability) 258.931 (0.092)
Source: Author’s own computation based on sample observation, 2012-13 *, ** and *** imply that coefficients are significant at level 1%, 5% and 10% respectively.
Let us now interpret the impact of selected community level traits on women’s
empowerment at the household level. A large section of the sample women have joined
SHG-centric microfinance programme which has nowadays taken the shape of social
190
movement. It inspired us to take the duration of SHG-membership as an explanatory
variable in the spectrum of community characteristics in the determination of women’s
empowerment at the household level. The coefficient of the duration of SHG
membership tells us that one year extra participation in SHG increases the empowerment
at the household level by 0.24% points. This result is statistically significant at 19% level
in Model-4A and 2% level in Model-4B. The logic behind this result comes out as
follows. First, the participation in SHG ensures the access to formal savings and credit of
the member women. It increases the financial asset holding of the women. Second,
access to formal credit increases the importance of the women within household. Third,
frequent meeting of group inculcate the political consciousness and democratic behavior
of the women. As a result, the duration of SHG membership has a direct effect on
women’s empowerment at the household level.
Caste is another community level trait which has been considered as a determinant of
women’s empowerment at the household level. This study shows that the coefficients of
the dummies for caste (SC=1) and caste (OBC=1) are positive. It points out that
empowerment at the household level is higher for the women under scheduled castes and
other backward classes in contrast to that for the women under general castes. On the
other hand, the coefficient of dummy for caste (ST=1) tells us that women belonging to
scheduled tribes are less empowered compared to women belonging to general castes.
However, Caste variables are not statistically significant determinants of women’s
empowerment at the household level.
Hence, estimates of women’s empowerment at the household level reveal that age,
occupation personal income, financial inclusion of the women, family type, dependency
ratio, household occupation, highest female education in the family are the major
determinants of women’s empowerment at the household level for the women in the
district of Bankura.
5.5.2. Determinants of Women’s Empowerment at the Community Level
We have computed the community level women’s empowerment applying two
alternative methodologies. The multiple regression models relating to the women’s
empowerment at the community level include some quantitative explanatory variables
and some dummy variables. So, these are known as Analysis of Covariance Models
191
(ANCOVA). In this sub-section we interpret the estimated ANCOVA models relating to
women’s empowerment at the community level. We have considered the same
household and community characteristics as explanatory variables in the model for
community level empowerment as they were in the model for household level
empowerment of women. First, we discuss the findings of the model (Model-4C) where
community level women’s empowerment has been measured by simple average of its
indicators. Next we explain the models (Model-4D) of community level women’s
empowerment measured by weighed average of principal components of the indicators.
The empirical estimates of Model-4C and Model-4D have been presented in table-5.5.3
and table-5.5.4 respectively. We find that goodness of fit for both the models is at
satisfactory level. F-statistic is statistically significant for both the models. The result of
White test of heteroscedasticity confirms the absence of heteroscedasticity problem in
the estimates of these models.
No doubt age of woman is an imperative factor for developing personality and mobility
in the society. The coefficient of age group (< 25 years) is positive. It tells us that the age
group (< 25 years) enjoys 0.82% point higher empowerment at the community level
compared to that of the aged group. But the result is statistically insignificant in both the
models. The age group AGE2 (25-35 years) of the women is found to be favourable for
community level empowerment. The average degree of empowerment at community
level of the women belonging to AGE2 group (belonging to age group AGE3 (36-45
Years)) is 3.7% (3%) points higher than that of AGE4. Further, in the estimation of
community level empowerment measured by PCA we have got qualitatively the same
result regarding the effect of age on women’s empowerment. Therefore, from this
finding we conclude that middle aged women in contrast to the aged women are more
empowered at the community level. Thus, middle aged women relative to aged women
have more mobility and socio economic consciousness.
The coefficients of the education level of women are positive and statistically significant
in Model-4C and in Model-4D. The coefficient of women’s education demonstrates that
one extra completed year of education increases the community level empowerment by
0.34% points from the mean level of empowerment. These results justify our hypotheses.
It is indicative that women’s education is a key factor in the determination of the
women’s empowerment at the community level in Bankura district. Although we have
192
reported that education of women is immaterial in the determination of household level
empowerment, but education has positive and significant impact on community level
empowerment of the women. Higher level of education expands the job opportunity,
mobility and accessibility to social institutions which are imperative to empowerment at
the community level. However, we should remember that most of the sample women
have education below the primary level. As a result we have got a marginal effect of
women’s education on their community level empowerment.
Among the dummies for personal occupations the coefficient of personal occupation
LAB (Wage labour=1) is negative and statistically significant. It implies that average
community level empowerment of wage labour women is 2.32% points lower than that
of the homemakers. Usually the earnings of women wage labourers are low and
controlled by their family members, particularly, by their husbands. As they are illiterate
or just literates and poor, they are less concern regarding the economic decisions and
several rights and opportunities. Thereby they have less participation on social and
community activities. Moreover, wage labour is not a respectful occupation in our
society. That is why, community level empowerment of the wage labourers is lower than
compared to that of homemakers. On the other hand, the coefficient of the dummy for
personal occupation SELF (Self employed or service=1) is positive but statistically
insignificant. It says that self-employed or service holder women are more empowered
than the homemakers in their community. This result has justified the expected direction
of relation. These findings are almost same for Model-4D.
In the estimation of the Model-4A and Model-4B we have seen that personal income of
women significantly affects women’s empowerment at the household level. But for
community level, the coefficients of personal income are positive but statistically
insignificant in Model-4C and Model-4D. Therefore, our empirical results explore that
personal income is important for accelerating household level empowerment but it is less
important to increase the degree of empowerment at the community level. We have
observed that most of the women earn from informal sector and they are ill paid.
Besides, these jobs do not get respect in our traditional society. This causes the
insignificant relation between the personal income and community level empowerment
of the women.
193
Table-5.5.3 Estimates of degree of Women’s Empowerment at Community Level
Dependent Variable: Degree of Women’s Empowerment at Community Level (DOWEC)
Method: Least Squares
Sample: 1 580
Included observations: 580
Explanatory Variable Coefficient Std. Error t-Statistic Prob.
Constant 41.023 1.985 20.670 0.000
Individual/Household Characteristics
Age Group (<25 Years) AGE1 (Yes=1) 0.828 1.673 0.495 0.621
Age Group (25-35 Years) AGE2 (Yes=1) 3.692*
1.255 2.942 0.003
Age Group (36-45 years) AGE3 (Yes=1) 3.047* 1.173 2.598 0.010
Education of the Woman, EDU (Year) 0.347**
0.160 2.163 0.031
Occupation of the Woman, LAB (Wage
Labour=1) -2.324
* 0.940 -2.472 0.014
Occupation of the Woman, SELF (Self employed/
Service =1) 2.106 1.556 1.354 0.176
Average Monthly Personal Income PINC (Rs. '00) 0.032 0.023 1.427 0.154
Access to Formal Credit AFCT (Yes=1) 7.457* 0.968 7.704 0.000
Type of Family TYFAMI (Nuclear =1) 2.286**
1.062 2.153 0.032
Dependency Ratio in the Household DRATIO (%) -0.064* 0.021 -3.065 0.002
Per Capita Household Income (APCHIN) (Rs.
‘000) 0.017 0.037 0.456 0.648
Household Occupation CULTI, (Cultivation =1)
-2.534**
1.077 -2.353 0.019
Household Occupation NONFARM (Non-Farm=
1) -0.403 1.378 -0.293 0.770
Household’s Land holding, HLAND, (bigha) 0.459* 0.166 2.775 0.006
Highest education among male household
members, HIMEDU,(Year) -0.008 0.116 -0.067 0.947
Highest education among female household
members HIFEDU(Year) 0.390
* 0.134 2.923 0.004
Community Level Characteristics
Duration of SHG-membership DSHGM (Month) 0.076* 0.014 5.542 0.000
Caste (OBC =1) 0.699 1.087 0.643 0.520
Caste (SC=1) 1.276 1.134 1.125 0.261
Caste (ST=1) -3.420**
1.514 -2.260 0.024
Summary Statistics
R-squared 0.535 Akaike information
criterion 7.340
Adjusted R-squared 0.518 Schwarz criterion 7.505
Durbin-Watson statistic 1.809 F-statistic (Probability) 30.609(0.000)
White Heteroskedasticity Test: H0: Variance of random disturbance term is constant
F-statistic (Probability) 1.065(0.298) Obs×R-squared (Probability) 239.167(0.325)
Source: Author’s own computation based on sample observations 2012-13
*, ** and *** imply that coefficients are significant at level 1%, 5% and 10% respectively.
194
Table-5.5.4 Estimates of Composite Women’s Empowerment Index at Community
Level
Dependent Variable: Composite Index of Women’s Empowerment at Community Level (CIWEC)
Method: Least Squares
Included observations: 580
Explanatory Variable Coefficient Std. Error t-Statistic Prob.
Constant -19.691 2.835 -6.947 0.000
Individual/Household Characteristics
Age Group (<25 Years) AGE1 (Yes=1) 0.306 2.389 0.128 0.898
Age Group (25-35 Years) AGE2 (Yes=1) 4.723*
1.792 2.635 0.009
Age Group (36-45 Years) AGE3 (Yes=1) 3.606**
1.675 2.153 0.032
Education of the Woman, EDU (Year) 0.693* 0.229 3.027 0.003
Occupation of the Woman, LAB (Wage Labour=1) -3.126**
1.343 -2.328 0.020
Occupation of the Woman, SELF (Self employed or
Service =1) 1.334 2.222 0.600 0.548
Average Monthly Personal Income PINC (Rs. '00) -0.008 0.032 -0.237 0.813
Access to Formal Credit AFCT (Yes=1) 11.937* 1.382 8.635 0.000
Type of Family TYFAMI (Nuclear =1) 3.063**
1.516 2.020 0.044
Dependency Ratio in the Household DRATIO (%) -0.039 0.030 -1.318 0.188
Per Capita Household Income (APCHIN) (Rs.
‘000) 0.065 0.053 1.235 0.218
Household Occupation CULTI, (Cultivation =1)
-2.563***
1.538 -1.667 0.096
Household Occupation NONFARM (Non-Farm= 1) -0.225 1.968 -0.114 0.909
Household’s Land holding, HLAND, (bigha) 0.551**
0.236 2.329 0.020
Highest education among male household
members, HIMEDU,(Year) -0.030 0.165 -0.184 0.854
Highest education among female household
members HIFEDU(Year) 0.503
* 0.191 2.635 0.009
Community Level Characteristics
Duration of SHG-membership DSHGM (Month) 0.078* 0.020 4.002 0.000
Caste (OBC =1) -0.775 1.552 -0.499 0.618
Caste (SC=1) -0.148 1.620 -0.092 0.927
Caste (ST=1) -8.596* 2.162 -3.976 0.000
Summary Statistics
R-squared 0.526 Akaike information criterion 8.053
Adjusted R-squared 0.508 Schwarz criterion 8.218
Durbin-Watson statistic 1.862 F-statistic (Probability) 29.487 (0.000)
White Heteroskedasticity Test: H0: Variance of random disturbance term is constant
F-statistic (Probability) 0.862(0.888) Obs×R-squared (Probability) 210.152 (0.822)
Source: Author’s own computation based on sample observations2012-13
*, ** and *** imply that coefficients are significant at level 1%, 5% and 10% respectively.
The coefficient of financial inclusion i.e. the coefficient of dummy AFCT (access to
formal credit, Yes =1) in Model-4C and Model-4D are found to be positive and highly
significant. The coefficient of access to formal credit indicates that the community level
195
empowerment of the women will increase by 7.45% points if she has access to formal
credit. When a woman can borrow form formal institution it makes an identity in society
or in social institutions. Again access to formal credit, particularly, from SHGs or
cooperative inculcates the banking habits and democratic idea allowing them in election
and selection process. In this way formal borrowing accelerates the women’s
empowerment at community level. Therefore, our study reveals that financial inclusion
increases women’s empowerment at the household level and at the community level.
It is seen that the type of family is a significant determinant of community level
empowerment of women. Our empirical estimate reports that community level
empowerment is 2.28% points higher of the women in nuclear families compared to the
women in joint/extended families. Women in the nuclear families enjoy more freedom to
participate in social and community decision making process. Besides, women in nuclear
families are compelled to move outside more for several household requirements which
definitely improve the community level empowerment. So, the result regarding the effect
of family type on community level empowerment of women is meaningful.
Like the estimation of the models for household level empowerment of women, we have
observed that dependency ratio has negative effect on indices of community level
empowerment. But it is statistically significant only in Model-4C. Table-5.5.3 shows that
one percent increase in dependency ratio from average level reduces the community
level empowerment of the women by 0.06% points from average. Although the finding
is supporting our hypothesis but the magnitude is negligible. Generally higher
dependency implies more number of older and children in the family. So, women spend
more time in the household’s jobs and rearing the children and older. It respects the
motherhood duties of women, but provides less time to manifest her potentially in
community. As a result women with higher dependency ratio have lower participation in
community level activities which are crucial for empowerment. So, our finding has a
logical base.
The empirical estimations of Model-4C and Model-4D confirm that per capita household
income increases women’s empowerment at the community level. However, the
196
empirical result is not statistically significant. Therefore, household income is immaterial
in the determination of women’s empowerment at the community level.
In table-5.5.3 the coefficient of the dummy for major household occupation
(Cultivation=1) tells us that the average degree of community level empowerment of the
women belonging to cultivator family is 2.53% points lower than that of the women
belonging to wage labour family. This result is statistically significant in both the
models. This finding goes against our hypothesis. But, this empirical finding is
meaningful. Usually the women of wage labour family are also wage labour. These
women come in contact with many people outside home. It helps them get experience in
broader arena of life. We also find that degree of empowerment at the community level
of the women under self employed family is lower than that of the women under wage
labour family. However, this result is not statistically significant at all.
Our empirical study reveals household’s landholding as a crucial factor for accelerating
women’s empowerment at the community level in the district of Bankura. The
coefficient of landholding in table-5.5.3 and in table-5.5.4 imply that one bigha extra
landholding increases the community level empowerment by 0.45% points and by 0.55%
respectively. Large land size of the household secures livelihood of the family and of the
women. Often women have the land ownership in the household with large landholdings.
These matters may enhance dignity of the women within and outside home and
accordingly the empowerment at the household and community level.
It has been obtained that the coefficients of highest education among the male household
members are statistically insignificant as an explanatory variable in the models for
women’s empowerment at the community level. However, the coefficient of the highest
education among the female members is positive and statistically significant at 1% level.
The coefficient implies that one extra completed year of education of the highest
qualified female member increases the average degree of empowerment at the
community level by 0.39% points in the area under study. We have got almost same
result in Model-4D regarding the highest female education. Therefore, female education
is more imperative than the male education in the family for inculcating women’s
empowerment at the community level. It is expected that most educated woman inspires
197
the other women to develop their personality in and outside home. Hence, our result is
significantly conclusive.
In the spectrum of community characteristics we have duration of SHG-membership and
caste variables as determinants of women’s empowerment at community level. In table-
5.5.3 and table-5.5.4 we find the coefficient of the duration of SHG-membership positive
and statistically significant. This result tells us that one year extra participation in SHG
increases the empowerment at the community level by 1% point. The logic behind this
outcome as follows. Participation in SHG ensures the financial inclusion allowing the
opportunity of formal savings, financial literacy and credit. These increase the financial
asset holding of the women and thereby the importance of the women within and outside
home. Further, frequent meeting of the group inculcate the political consciousness and
the association among members reduces several social barriers of empowerment.
Therefore, it is not surprising that SHG-membership induces women’s empowerment at
the community level.
This study shows that the coefficients of the dummies for caste (SC=1) and caste
(OBC=1) are positive. It may be argued that empowerment at the community level is
higher for the women under scheduled castes and other backward classes in contrast to
that for the women under general castes. However, we are not worried of these results, as
they are not statistically significant at all. On the other hand, the coefficient of dummy
for caste (ST=1) is negative and statistically significant. Women belonging to scheduled
tribes are less empowered at the community level compared to women belonging to
general castes. We observed that due to language barriers and community rituals tribal
women are socially backward in the context of general castes. Tribal women are less
educated compared to general caste women. As a result, their empowerment in respect to
other castes is lying far below. However, it should be noted that tribal women have full
dignity in their own society.
We, therefore, conclude from this study that age, education, occupation, financial
inclusion of the women, family type, dependency ratio, household occupation, highest
female education in the family, duration of participation in SHG and caste are the major
determinants of community level empowerment for the women in the district of Bankura
West Bengal.
198
5.6. Conclusion
This chapter has vividly discussed the empirical finding of our estimated econometric
models. We conclude this chapter with the significantly conclusive findings of the study
as follows. We find that women’s empowerment at the household level and at the
community level significantly increases the probability of adopting family planning.
Household level empowerment is more important than community level empowerment
of women for reducing the probability of experiencing domestic violence. We have
obtained that both the empowerment has some favourable impact on child education
expenditure. Therefore, women’s empowerment is an effective factor for improving
household welfare. Finally, our study has shown that women’s age, personal income,
personal occupation, financial inclusion, family type, household occupation and
participation in SHG are major determinants of household level empowerment. On the
other hand, women’s education, personal occupation, family type, household occupation,
highest female education and duration of SHG-membership are important determinants
of community level empowerment. Based on the empirical findings, we may suggest
some policies for improving the level of empowerment and its consequences in
households and in society at least in the area under study. In chapter six we suggest the
alternative policy prescriptions on the basis of our empirical results.
199
Chapter Six _______________________________________________________________________
POLICY PRESCRIPTIONS
6.1. Introduction
In order to explore a glimpse of women’s empowerment in Bankura district we have
computed women’s empowerment at the household level and at the community level
using a set of primary data. In the last two chapters we have discussed the empirical
findings of our dissertation. On the basis of these empirical findings, we would like to
discuss the some policy issues relating to women’s empowerment in this chapter. This
chapter has been divided into five more sections. The major findings of this empirical
study have been reproduced in section 6.2. There are several government and non-
government programs in favour of women in India. We present them in section 6.3 along
with its sub-sections. In section 6.4 we have discussed the policy prescriptions and
implications based on our empirical findings. It has four sub-sections. We conclude the
study in section 6.5.
6.2. Major Findings
Let us recapitulate the major findings of our empirical study which help us formulate the
relevant policies towards better empowerment of women and its consequences in the
district of Bankura and in major rural area in India.
1) Majority of our sample households are poor. This study has shown that forty percent
of sample women have not adopted any family planning measure. Domestic violence
against women is a major socio-economic problem in the rural areas of Bankura district.
200
2) The average empowerment at the household level is higher than that at the community
level for the women in the district of Bankura. Only one third of the sample women in
our study are relatively empowered at the household level and at the community level.
3) Both women’s empowerment at the household level and empowerment at the
community level are positively correlated with the decision regarding family planning in
the area under study. The empowerment variables are negatively correlated with the
incidence of domestic violence against women. Child education expenditure as
proportion to the annual household income has a positive and statistically significant
correlation with mother’s empowerment.
4) The important findings of our empirical estimation of the decision regarding family
planning are as follows.
Household level empowerment of women directly affects the probability of
adopting family planning decision.
Households in the district of Bankura have bias for male child which reduce the
probability of taking family planning decision.
Age of woman at marriage and spousal age gap increase the probability towards
family planning.
Education of the woman and their husband are favourable for adopting family
planning decision.
The women belonging to non-farm self employed family are less likely to adopt
family planning decision.
The dependency ratio in the family adversely affects the probability towards
adopting family planning decision.
Household income is an important determinant of the decision regarding family
planning.
The community level empowerment is instrumental for adopting family planning
decision.
Participation towards SHG-centric microfinance program persuades the women to
take family planning decision.
Tribal women in contrast to the women under general castes are less likely to take
family planning decision.
201
5) The main findings relating to the estimation of the probability of the incidence of
domestic violence against women are presented below.
Household level empowerment of women reduces the probability of facing
domestic violence of the women.
Education level of husband and other male members in the household reduces
domestic violence against women.
Women belonging to non-farm self-employment household are less victimised in
domestic violence than the rate of domestic violence of the women belonging to
wage labour class.
Higher the size of landholding of the household, the higher is the probability of
facing domestic violence for the women in the district of Bankura.
Household income is favourable to reduce the probability of domestic violence
against women.
Although household level empowerment significantly reduces the extent of
domestic violence, community level empowerment is less important for
combating the curse of domestic violence in the area under study.
Longer duration of SHG-membership increases the probability of facing domestic
violence against women.
Women from the scheduled castes and scheduled tribes in contrast to those from
general castes women are suffering more from domestic violence.
6) We now mention the important results regarding the issue of the proportion of
household income spent for child education.
Women’s empowerment at the household level successfully increases the share of
household income for child education for the rural households.
Our study reveals that father’s education adversely affects the child education
expenditure as proportion to household income.
Highest male education and highest female education have some favourable
impact on child education expenditure as proportion to household income.
The share of household income spent for child education is higher in nuclear
families than that in joint or extended families.
Dependency ratio positively affects the child education expenditure.
Higher per capita income reduces the share of spending for child education.
202
The community level empowerment of women increases the share of household
income for her child education.
Households in different castes in Bankura district are indifferent in terms of child
education expenditure.
7) We now turn to the major empirical determinants of women’s empowerment at the
household level.
The women under young age group enjoy higher empowerment in their
households than the women under older age group.
In contrast to homemakers, wage labour women enjoy lower but self-employed or
service holders enjoy higher level of household empowerment.
Though personal income of the women significantly increases the degree of
women’s empowerment at the household level, household income is less
important in this regard.
Financial inclusion has a favourable effect on household level empowerment.
The women of nuclear family enjoy higher empowerment within household
relative to the women of joint family.
The dependency ratio in the household has a negative impact on the degree of
women’s empowerment at the household level.
Household level empowerment of the women under cultivator households is lower
than that of the women under wage labour households.
It is interesting to mention that although women’s education has no significant
effect on their household level empowerment, highest female education
accelerates the household level empowerment of the women.
The duration of SHG-membership increases the household level empowerment.
The household level empowerment of the women doesn’t vary across the castes.
8) The major findings of community level empowerment are pointed below.
Community level empowerment, of the women under age group 25-35 years and
age group 36-45 years, are higher than that for the women of older age group.
Women’s education improves the community level empowerment of the rural
women. But, it is unimportant for determining household level empowerment.
203
Community level empowerment of the wage labour women is lower than that of
the homemakers.
The financial inclusion enhances women’s empowerment at the community level
in the area under study.
Compared to the women living in joint families the women living in nuclear
family enjoy higher empowerment in her community.
The dependency ratio in the household adversely affects the degree of women’s
empowerment at the community level.
Community level empowerment of the women under cultivator households is
lower than that of the women under wage labour households.
Household landholding directly affects the community level empowerment of the
rural women.
Highest female education in household significantly accelerates the community
level empowerment of the women.
Participation in SHG-based microfinance movement has inculcated the
community level empowerment of the sample women.
Community level empowerment of the tribal women is comparatively lower than
that of the general caste women.
6.3. Existing Policies and Programs towards Empowering Women
The Indian Constitution ensures the principle of gender equality. Moreover, it authorizes
the State to adopt measures of positive discrimination in favour of women for balancing
the cumulative social, economical, and political backwardness of women. Furthermore,
India is a signatory member of the Convention on Elimination of All Forms of
Discrimination against Women (CEDAW) in 1993. So India has taken several measures
for women to ensure gender equality in all spheres of life. Before going to the imperative
policy implications of our empirical study we review the existing policies and
legislations regarding women’s empowerment and related issues.
6.3.1. National Policy and Legislation for Women in India
The government of India has established the National Commission for Women in 1992
for monitoring all constitutional and legal security measures for women and to review
the existing legislations related to the rights of women so that women can enjoy equality
204
in all spheres of life and have equal participation in the development of the nation. It was
constituted as an apex level statutory body under the National Commission for Women
Act, 1990.
The 73rd
Constitutional Amendment Acts, 1992 ensure one-third of the total seats for
women in all elected offices in local bodies whether in rural areas or urban areas.
The National Plan of Action for the Girl Child (1991-2000) was taken to ensure survival,
protection and development of the girl child for better future.
In order to address women's advancement, development and empowerment, The
National Policy for Empowerment of Women, (NPEW) has been formulated in 2001.
The major objectives of this national policy were as follows
The advancement, development and empowerment of women in all spheres of life
Introduction of more responsive judicial systems in favour of women’s needs
Ensuring women's equality in power sharing and active participation in decision
making
Mainstreaming a gender perspective in development process
Strengthening and formation of relevant institutional mechanism
Partnership with community based organizations; and
Implementation of international obligations, commitments and cooperation at the
international, regional and sub-regional level
In February 2012, the Government of India has also formed a High Level Committee to
undertake comprehensive study to understand the status of women since 1989 as well to
evolve appropriate policy interventions based on a contemporary assessment of women's
needs. The committee is already functioning.
We have seen that recent policies and programmes of the Government for women
welfare are already directed towards achieving inclusive growth and in line with the
objectives of the national policy for women’s empowerment. Following Constitutional
mandate the state has passed several legislation for removing social discrimination
205
against women and the protection of the women from different social and cultural evils
against women. We now list some relevant legislation relating to our study.
The Dowry Prohibition Act, 1961
In order to protect the female from the curse of dowry and to prohibit the evil practice of
giving and taking of dowry our government has passed The Dowry Prohibition Act in
1961. The Act has been in implementation since 1961. The Act underwent amendments
in the year 1984 and again in 1986. The Act goes for further amendment for considering
the suggestions of National Commission for Women (2009). For this purpose, a Review
committee was constituted in 2012. In a meeting with the women Members of
Parliament in 2012 Ministry of Women and Child Development has suggested many
proposals for the amendments of the Dowry Prohibition Act. These are to be examined
and finalised in the Ministry.
The Indecent Representation of Women Act, 1986
The Indecent Representation of Women Act was enacted in 1986. The objective of this
act was to prohibit indecent representation of women in advertisement, publication,
writing and painting or in any other manner and references that are insulting to the
dignity of women. Violation of this Act is punishable with imprisonment up to two
years. Still now this Act is applicable only to the print media. Very recently, the
Government has approved amendments to apply the law to audio visual media and
material in electronic form and revising the penalties. The amendment Bill is currently
with the Parliament.
The Protection of Women from Domestic Violence Act (PWDVA), 2005
The PWDVA is a civil law in India. It favours the victim women of domestic violence
to get immediate support in the form of shelter, medical facility and reliefs in the nature
of protection, residence, compensation, maintenance and give orders for temporary
custody of children. The law also widens the meaning of the word 'aggrieved woman' by
including women who face domestic violence in relationships other than matrimonial
relationships like daughters, mothers, sisters and those involved in marriage like
relationships and providing a woman's right to reside in the shared household. The Act
came into force on 26th
October 2006. The Ministry has formulated a scheme for
206
assistance to State Governments for implementation of the Act, as a component of the
Umbrella scheme on Protection and Empowerment of Women.
Sexual Harassment of Women at Workplace (Prevention, Prohibition and Redressal),
Act 2013
The Act seeks to cover all women, irrespective of their age or employment status and
protect them from sexual harassment at workplaces both in public and private sector,
whether organised or unorganised. Women who are employed as well as those who enter
the workplaces as clients, customers and apprentices, students in educational institutions
and patients in hospitals etc. are also sought to be covered under this Act. Under this Act
Sexual Harassment of Women at Workplace is illegal and punishable.
Therefore, governments have taken several legal and constitutional steps for
development of the status of women. In addition to these constitutional and legal steps,
the governments have introduced several schemes and programme for improving overall
status of women.
6.3.2 Governmental Programmes for Enhancing Women’s Empowerment in India
Since independence the Government of India has been enacting different pro-female
laws and rules and implementing various plans and programmes to improve the status of
women in our country. In the First Five Year Plan period (1951-56), mainly welfare
oriented plans and programmes were taken for women. In that era the Central Social
Welfare Board (CSWB) undertook a number of welfare measures through the voluntary
sector. As far as women’s issues were concerned, the then government implemented the
programmes through the National Extension Service Programmes through Community
Development Blocks. In order to ensure better implementation of the welfare schemes
during the Second Five Year Plan period (1956-61) the Government emphasised to
organise “Mahila Mandals” (women’s groups) at grass-roots levels. Subsequently,
during the next three Five Years plans and four annual plans period, the governments put
importance on women’s education, improved maternal and child health services, feeding
for children, nursing and expectant mothers for improving the status of women. Sixth
Five Year Plan (1980-85) is regarded as a landmark in women’s development
programmes. In this Plan the government adopted a multidisciplinary approach with a
three-pronged thrust on health, education and employment of women. Development
207
programmes for women were continued with the next Plan. One of the major objectives
of Seventh Five Year Plan (1985-90) was the improvement of economic and social status
of women and brings them under the mainstream development process. In this plan
period an important step was to identify and promote “beneficiary-oriented programmes”
which extended direct benefits to women.
For women development, agendum of Eighth Five Year Plan period (1992-97) was to
ensure that the benefits of development of different sectors did not deprive women. For
this purpose some special programmes were implemented to complement the general
development programmes. In order to ensure the functions of women as equal partners
and participants in the developmental process the government introduced reservation in
the membership of local bodies. This unique initiative marks a sharp departure from
‘development’ to ‘empowerment’ of women. This initiatives of women’s empowerment
were forwarded to Ninth Five Year Plan (1997-2002) by adopting women’s component
plan at the Central and State levels. The Tenth Five Year Plan (2002-2007) emphasised
to ensure the requisite access to information, resources and services for women, and
advance gender equality goals. In order to achieve inclusive growth in the Eleventh Five
Year Plan (2007-2012) the authority had taken several initiatives which facilitate the
women to develop their full potential and share the benefit of economic growth and
prosperity. In this time government undertook special measures for gender
empowerment and equity. The Ministry of Women and Child Development has tried to
make synergistic use of gender budget and gender mainstreaming process. The Twelfth
Five Year Plan (2012-17) has emphasised on gender equity.
Let us now look at the brief history of the schemes and programmes which have been
framed to alleviate poverty vis-à-vis to increase the empowerment level of women. Some
of these programmes are IRDP, DWRCA, TRYSEM, NRDP, RLEGP, JRY, and SSEGS
etc. Different studies suggest that most of the programmes help the rural people little and
they are not self-sustaining (Swaminathan, 1990, Kaladhar, 1997). Moreover, if there
was any success of these programmes, it was in favour of men not in favour of women.
That means these programmes were not fruitful in generating women’s empowerment.
Under this background, in 1999, SHG was introduced under SGSY to improve socio-
economic conditions of women in general and empowerment in particular. We have
already said that the women are resource-poor and so they have no collateral assets or
208
security, which ensure them to have loan. SHG was mainly formed to provide formal
loan to the un-banked women of the society without physical collateral security. SHGs
not only accelerate the accessibility to formal credit for the member women but also
inculcate the values and dignity of the women in the society through generating their
empowerment. Therefore, the recent government policy has also emphasized on the
women’s empowerment in India.
In 2001, Swayamsiddha Scheme was introduced in West Bengal along with a few other
states in India as an additional program of the Integrated Child Development Scheme
(ICDS). It is a central Government sponsored program for empowering rural poor
women, economically and socially, through formation and mobilization of SHGs. The
project has been functioning in the district of Bankura since 2002 under the ICDS
network. In this scheme ‘anganwari’ workers facilitate the poor women to form and
nurture the SHGs. Finally, the ‘anganwari’ worker introduces the formed group with the
Banks. As per annual report 2005, in different states the SHG members under the
scheme are doing income generating activities such as food preservation, dairy farming,
cutting and tailoring, embroidery, kitchen gardening, beautician, rope making, etc.
However, in our study area, no SHG under this scheme undertakes any income-
generating activity. This scheme is not implemented in all the states of India. Recently
the government of India has stopped the funding for this scheme. As a result the existing
groups have also lost their power of functioning. So this type of short run scheme is not
suitable for enhancing the empowerment of the rural women.
Currently, the following schemes and programmes have been functioning under the
Ministry of Women and Child Development for assisting the women and children in
India (Statistics on Women in India, 2010).
Support to Training and Employment Programme (STEP)
Rajiv Gandhi Scheme for Empowerment of Adolescent Girls, SABLA
Swawlamban
Construction/Expansion of Hostel Building for Working Women with a Day Care
Centre (WWH)
Balika Samriddhi Yojana (BSY)
209
National Programme for Adolescent Girls (Kishori Shakti Yojana)
Shishu Greha Scheme
Integrated Scheme for Street Children
Scheme for Welfare of Working Children in Need of Care and Protection
Prevention and Control of Juvenile Maladjustment
Integrated Child Protection Scheme (ICPS)
Conditional Cash Transfer Scheme for the Girl Child with Insurance Cover
General Grant-in-Aid for Voluntary Organisations in the Field of Women and
Child Development
National Mission of Empowerment of Women
Scheme for Leadership Development of Minority Women
Conditional Maternity Benefit Scheme
Education Scheme, Food and Nutrition Board (FNB)
Ujjawala, Scheme for Combating Trafficking
Nutrition Programme for Adolescent Girls (NPAG)
Wheat Based Nutrition Programme
Anganwadi Karyakati Bima Yojana
Therefore, governments have taken a lot of initiatives for improving the socio-economic
and demographic status of women in India. To meet the needs of women and children
there has been a progressive increase in the plan outlays over the last six decades of
planned development. The outlay of Rs. 4 crores in the First Plan (1951-56) has
increased to Rs. 13,780 crores in the Tenth Five Year Plan. The important feature of the
programmes for women was “welfare” oriented during the pre liberalisation era (1951-
1979). During the weak liberalisation era (1980-1990) these were “development”
oriented. There has been a shift from “development” oriented approach in the weak
liberalisation era to “empowerment” of women in strong liberalisation era (since1991).
In 2010, the Government of India has taken National Mission for Empowerment of
Women (NMEW). It is an initiative for holistic empowerment of women by securing
convergence of schemes/ programmes of different Ministries/Department of Central
Government as well as State Governments. In addition to these schemes the governments
have conducted several conferences, seminar and events relating to women’s
empowerment during the last few years. Very recently, the Government of West Bengal
210
has taken ‘Kanyashree Prakalpa’. It is expected to be fruitful to improve the status and
well being of the girl child in West Bengal by incentivizing schooling of all teenage girls
and delaying their marriages until the age of 18, the legal age of marriage.
6.4. Policy Prescriptions and Implications
This empirical study has measured the empowerment of women at the household level
and at the community level in Bankura district of West Bengal using two alternative
methodologies. We have assessed the impact of women’s empowerment along with
some socio-economic and demographic characteristics on the decision regarding family
planning, on incidence of domestic violence against women and on children’s education
expenditure. It has also identified the factors responsible for affecting empowerments.
On the basis of these empirical findings, we can suggest some supplementary and
alterative policies for empowering women. These policies may help the policy makers to
inculcate women’s empowerment at the household level and at the community level in a
better way in future. We are going to discuss the policies, which come out as a
consequence of this empirical study conducted in the district of Bankura.
Our field observation reveals that a large portion of our sample women have very little
empowerment at the household level and at the community level. It indicates that the
existing policies are not enough for the improvement of the empowerment among
women. Therefore, inculcation of women’s empowerment is a challenge ahead to the
government and policy makers in our country. In spite of this, we find that women’s
empowerment at the household level and at the community level significantly increases
the probability of adopting family planning program and reduces probability of domestic
violence against her. Further, women’s empowerment at the household level and at the
community level in Bankura district have improved their spending attitudes towards
child education. It is indicative that women’s empowerment are favourable for
household and child welfare, broadly social welfare. Therefore, in order to increase the
welfare of our study area we can suggest for improving empowerment of women in the
district of Bankura. The particular policies which are appropriate for improving women’s
empowerment have been proposed in section 6.4.1.
211
6.4.1. Suggestions for Empowering Women in Bankura District
In our study we have established that women’s empowerment at the household level and
at the community level increases the probability of adopting family planning and reduces
the probability of facing domestic violence. Community level empowerment increases
the share of household expenditure on child education. Therefore, women empowerment
at the household level and at the community level is an alternative instrument for
improving household and child welfare in the district of Bankura. Hence, we need to
search the factors suitable for enhancing women’s empowerment at the household level
and at the community level. We have done this job in the estimation of the models for
women’s empowerment. Based on the empirical results we recommend the following
policy towards empowering women at their households and community.
First, this dissertation concludes that elder women in the district of Bankura have lower
level of empowerment relative to the women in the younger age groups. We think that it
happens due to livelihood insecurity and lower control over household assets. In order to
change this picture we need to have some special package for empowering elder women
and for ensuring livelihood security of the older women. In this regards NGOs like
DHAN Foundation in Tamil Nadu, SEWA in Gujarat, function effectively. SEWA
extends social security to its members by ensuring the access to health care, shelter,
banking services and income security. On the other hand DHAN foundation provides
mutual insurance called ‘People Mutuals’ for social security. This initiative safeguards
the poor from risk and vulnerabilities through mutual solutions and collaboration with
insurance providers. In addition to our existing insurance policies we, therefore, propose
to launch these types of policies for livelihood security of the elder women in rural
Bankura. In this case local panchayet, existing women organization, NGOs would be
more helpful.
Second, women’s education is very much important to improve the community level
empowerment of women. In addition to this we have found that highest education level
of the female members in the households is favourable factor for empowering women.
Education has an intergenerational impact on women’s empowerment. However, in our
study we have observed that most of the sample women cannot cross the primary
education level. All women should be ensured of getting minimum level of education. It
is, therefore, a common policy demand that we have to increase the level of women’s
212
education for improving their empowerment. So, it is necessary to setup more
educational institution for women. It is also necessary to monitor the enrolment of
women in the educational institution. We can also suggest for establishing informal
educational institutions by NGOs or other voluntary organizations for aged women.
Besides, the government should implement women’s education programme more
effectively in rural areas, particularly in the area under study.
Third, self-employed and service holder women have higher empowerment at the
household level and at the community level compared to the empowerment of the
homemakers. Further, personal income is imperative for enhancing empowerment at the
household level and at the community level. So we have to create an environment that
helps the women participate in formal workforce or participate in self-employment
activity. We have seen that, in Bankura district, there are some famous cottage industries
like ‘Teracota’, ‘Dogra’, ‘Baluchari Silk’, ‘Bishnupuri silk’ and ‘cotton Handloom’.
These industries are dominated by women workers. But the fact is that these women
worker are still informal and unskilled in nature and thereby controlled by male persons
and to some extent by moneylenders. So the government should implement financial
assistance and different training programmes that will be helpful to make women
financially independent and to develop skill among women. In Gujarat SEWA, a
membership based organization of self employed women, has been playing a
commendable role in the empowerment of women. It works to ensure full employment
and self-reliance for its members. We may propose to establish the organization like
SEWA to extend the self employed activities and hence self-reliance of the women in the
district of Bankura. Not only that we have to be conscious regarding the formal
employability of the educated women. All these will be helpful to achieve the goal of
employability and increased income of women effectively. Thereby empowerment of
women will increase automatically.
Fourth, it has been found that access to formal credit is an important factor to inculcate
the empowerment of women. Therefore, banking institutions and non-bank financial
institutions have to take some special programs for ensuring the access to formal credit
for the rural women. In this regards bank have to relax some conditions like asset
holding, income earning that hinder the accessibility of formal credit for the rural
women. Besides, banks should expand their branches in rural area.
213
Fifth, the duration of SHG-membership increases the community level empowerment of
the member women. So, we can say that the SHG-based microfinance program plays an
active role to accelerate community level empowerment of the women. Therefore, we
have to implement the SGSY policy in more intensive and extensive way and we have to
encourage the NGOs with some incentives. Formal financial institutions have to be more
enthusiastic in the matter of financial inclusion of the rural women.
Sixth, this study has reported that women belonging to scheduled tribes are less
empowered at the community level contrasted to the women belonging to general caste
women. During the time of data collection we have observed that ST people have very
poor accessibility to various primary needs of life like education, health, housing etc.
They do not come forward to participate in community activities. These lead to have
lowering effect on empowerment of ST people. So in order to improve the community
level empowerment of the tribal women we need to take some exclusive programs for
ST women like development of education system by tribal language, social awareness
programs etc. Some special programme should be taken to increase the confidence of
tribal women so that they can come forward to organize social or community
development programme.
From our field observations show that most of the sample women do not know about the
legal rights and government initiatives in favour of them. They live within the periphery
of social, cultural, regional and caste customs. They are unaware of their right. These are
the causes of low empowerment of women of the selected region under study.
Government should establish a legal cell integrated with the panchayet to give free
advice to women about their rights. Again community leaders and local elite should
serve and support women in social participation. They can jointly organize meeting,
group discussion, speeches and counseling for motivating people about gender equality
and encouraging women to participate in community development. Moreover various
programmes in electronic media, organizing rallies and public meeting could be effective
tools to create massive awareness among women and thereby help in empowering of
women.
Women, as we find in the district of Bankura during the time of field survey, are oriented
internally in such ways that they think that living under the layer of someone like father,
214
husband and son is respectable for them. They also believe that women should obey the
instructions of their husband, sons, father etc. regarding family and social decision. They
should not inherit physical asset etc. This type of internal orientation of the women is
really a constraint of women’s empowerment. So to bring to the change in the internal
orientation of women, long-term consciousness generation programmes are badly needed
in the area under study. Governmemt should think over it and do the needful.
Finally, we know that some leading international organizations such as World Bank,
WHO, UNICEF, UNDP etc. launch various programmes for the development of women
in India by funding in different projects related to education, health, human rights etc.
They have been playing a major role in eliminating gender discrimination. These are
implemented through government organization, NGOs and women’s organizations. But
in the region of our study these organizations are not functioning well. So the
government should take some steps so that various local organizations can support the
international organizations to be more effective.
6.4.2 Suggestions for Improving the Likelihood towards Family Planning
In the estimated model for the decision regarding family planning we have found that in
addition to women’s empowerment there are several socio-economic-demographic
features which significantly affect the decision regarding family planning. Based on
these results, we can prescribe the policies for improving the likelihood towards family
planning decision.
First, it has been reported that women’s empowerment at the household level and at the
community level significantly improved the likelihood towards family planning. So, we
have to repair the loopholes of the existing policies/schemes for improving women’s
empowerment at the household level and at the community level. In section 6.4.1 we
have already suggested some alternative policies which are suitable for enhancing
empowerment of women in the area under study.
Second, our study has shown that there is male child bias which is a crucial impediment
towards taking family planning decision in Bankura district. In order to neutralise this
bias we need to inculcate awareness regarding gender equality. So governments,
particularly the health departments, and NGOs have to launch the awareness generation
215
programs in the rural areas in more and more extensive form. We, the members of civil
society, also have to be dutiful to guide our fellow citizens regarding gender equality
and the importance of family planning. Lesson of gender equality should be
incorporated in the school level syllabus. It has been turned out that age at marriage
directly affects the probability towards family planning decision. Further spousal age
gap is suitable for family planning. We have to restrict child marriage. On the other
hand, we have to allow age gap between bride and groom. We have already legal age of
marriage (18 years for female and 21 years for male) but in rural area majority of the
parents are not conscious of the law and regulations and its implications. The
‘Kanyashree Prakalpa’ of the Government of West Bengal is no doubt an innovative
step to stop female child marriage and to encourage female child education. Besides, we
have to play a vital role to aware them about the bad effect of child marriage. Health
department should arrange various programmes about the bad effect of lower age
pregnancy and frequent pregnancy and its effect on women’s health.
Third, we find that the level of education of women is an important factor of the
probability of taking family planning decision. Husband’s education has also a positive
effect on the probability of adopting family planning decision. However, the sample
women and their respective husbands do not have the significant level of education.
Educated parents are eager to have healthy baby rather than have large number of
babies. They want to spend more for their children’s health, education. If they have large
number of babies it becomes very difficult for parents to spend more on their children’s
education and health. Therefore, the government and the social institutions have to take
incentives to extend the education facilities. In addition to the extension of formal
education the government should arrange health conscious training, awareness
generation programme etc. in order to make the family planning programs successful.
Government can open women education center, girls’ schools, adult education center in
the rural area. Again Government can provide subsidies and different facilities to spread
women education.
Fourth, our study shows that wage employed class where women are also employed have
higher probability of taking family planning decision than farm or self-employed
occupation and higher dependency ratio reduces the probability of taking family
planning decision. So, higher worker population ratio in the household is suitable for
216
taking the decision family planning. Therefore, we need to create an environment that
helps people to move from farm or self-employed occupation to wage employment
occupation for improving the status of family planning in our study area. The
government needs to consider industrialization exclusively for this district or favours
private companies in this regard. Besides, it is needed to improve the communication
system in the district to help the people for searching and holding suitable wage
employment.
Fifth, household income is directly related with the probability of taking family planning
decision. So, it is needed to strengthen the policies of income poverty alleviation and
policies regarding employment generation of the rural households. In this sense NREGA
is suitable one. Besides, the government may take some policies in favour of small and
cottage industries and agro-based industrialization.
Sixth, we find an encouraging effect of SHG centric microfinance program on the
likelihood of adopting family planning decision. This finding establishes the
effectiveness of the SGSY program in household welfare. Therefore, we have to
implement this group based microfinance program more actively and extend it through
opening more and more channels in the rural areas.
Seventh, our study concludes that the scheduled tribe women who are the most deprived
section in the district are less likely to take family planning decision in contrast to the
other castes. They have deep dependence on their own social customs. During the time
of field survey, we have observed that the people of tribal community are most ignorant
of the government policies and health facilities available for them. Most of them have no
formal education or vocational training. Although there exists some special program like
LAMPS (Large Sized Agricultural Multi-purpose Cooperatives Society) for economic
development of the tribal community but the result of our study indicates the
insufficiency of these programs. Therefore, government has to launch some exclusive
health care program for this community for inculcating family planning habits.
6.4.3. Suggestions for Alleviating Domestic Violence against Women
Violence against women is a serious problem in our society. The present data regarding
violence against women inside and outside home tell us that existing polices failed to
217
save women from violence. We now prescribe the alternative policies for reducing the
likelihood of domestic violence against women.
First, we have found that women’s empowerment at the household level is necessary for
curbing domestic violence against women. It indicates that we have to ensure her
empowerment inside home. For this purpose, we have to rectify the customs and
resolution which give space for the women in the socio-economic decision making
process in their households. Besides empowerment at the household level, we have
identified several socio-economic-demographic traits influencing the probability of
facing domestic violence.
Second, our empirical estimation reveals husband education as a panacea for reducing
likelihood of domestic violence against women. Not only that, education of other male
members is important to reduce the suffering of women from domestic violence. But we
have seen that average education level of husbands of our sample women is just at
primary level and average education of other male persons in the households is less than
eighth standard. Therefore, the expansion of male education is urgent in order to prevent
the women from the disease of domestic violence. For this purpose, we have to spread
our educational system. That means initiatives should be taken to increase the qualitative
as well as quantitative aspects of education.
Third, the nature of household occupation, Cultivation and Non-farm self employment,
reduce the probability of domestic violence. Further, the household landholding
increases the probability of domestic violence. We also observe that even in landed
households women do not have any land ownership. It makes gender inequality and
sometimes, causes violence against female members of the family. In order to reduce the
curse of domestic violence against women, redistribution of land in favour of women is
required. We agree that in India, particularly in our state of West Bengal, land reforms
have been progressed during the last fifty years towards landless farmers. But it is hard
to find out the programs and policies towards land redistribution in favour of women in
our state and in the area under study. This negligence may be a vital cause of the low
status of women. Therefore, for reducing domestic violence we recommend to take some
effective land reform policies and programs which make redistribution of land in favour
of women. We may follow the land purchase scheme for SC/ST women in Tamil Nadu.
218
Under this scheme, landless women can purchase land for cultivation with a maximum
project cost rupees two lakhs. This scheme entails 50 per cent subsidy from Tamil Nadu
Adi Dravidar Housing and Development Corporation Ltd. and remaining part comes as
bank loan.
Fourth, household income is inversely related with the probability of facing domestic
violence. We, therefore, have to strengthen the policies of income and employment
generation for the rural households. Government may take some policies in favour of
small and cottage industries and agro-based industrialization for rural households which
are indirectly helpful to reduce domestic violence.
Fifth, this study has reported that dowry at marriage and at post marriage amplifies the
likelihood of domestic violence against women. We have found that drug addiction of
the husband is a major cause of domestic violence. So dowry deterrence act and laws and
regulation against drug addiction have to be implemented seriously. In addition to the
acts and regulations we have to campaign in favour of dowry deterrence and against drug
addiction. We need to inculcate the consciousness among people about the curse of
dowry and drug. In this regard Governmental officials and NGOs, and even we the
common people may take part in the conscious generation programs.
Sixth, our study shows that the scheduled tribe women are more victimized compared to
other sample women. Although there exists some dedicated program like LAMPS for
economic development of the tribal community but the result of our study indicates the
insufficiency of these programs. Therefore, government has to launch some exclusive
consciousness program for this community for reducing domestic violence in this
community.
6.4.4. Suggestions for Improving Children’s Education Expenditure
In this section we have proposed some policies for improving child education
expenditure of the households in the district of Bankura.
First, women’s empowerment at the community level has been found as an important
factor affecting child education expenditure out of household income. So improvement
of community level empowerment of women is an alternative approach for improving
219
the proportion of household expenditure on child education. Suggestions for
improvement of women’s empowerment at the community level have already been
presented in section 6.4.1.
Second, we find highest education of male and female in the family as a stimulating
factor for increasing children’s education expenditure. In other words, the household
with low level of educational background, as we have observed in the course of field
survey, do not like to spend more for children’s education and thereby the children
remain uneducated which in future will be the cause of low education of future children
of the household. Therefore, low education of the household members makes an inter-
generational vicious circle of education. Therefore, in order to break this vicious circle,
the government needs to implement some policies to ensure the access to higher
education for the rural people at affordable cost. Nowadays government has already
taken some policies to expand the higher education across our country. But the point is
that the government needs to monitor the existing system and implement new institutions
making clear-cut vision and mission for better achievement.
Third, we have observed that poor household could not spend the necessary amount for
their children’s education. Although, occupation is insignificant in the determination of
children’s education expenditure, we find that income is vital for it. So we suggest for
implementing more income generating projects in rural area as like NREGA and to some
extent SGSY etc.
Fourth, children’s education expenditure as proportion to household income of lower
caste households like OBC, SC and ST is lower than that of the general caste
households. So we need to have some special financial support programs for the lower
caste households regarding children’s education. We know that for the students
belonging SC and ST financial support system already exists. But these are not sufficient
and have very low coverage at the initial stage of education. However, for the poor
students of OBC and General caste, there is no such financial support system. Therefore,
we propose to the governments and private institutions to provide a range of subsidy in
fees or financial support for poor family for education of their children who are the
future our nation also.
220
Thus, we can conclude that to achieve the goal of real development we have to create an
environment where women get equal dignified opportunity to work hand in hand with
men. The policies recommended above actually are the results of our empirical research
conducted in the district of Bankura. It is, therefore, obvious that these policies are
suitable for the women residing this district. These policies may also be applicable for
the people of any region with the same type of socio-economic and demographic
characteristics as in Bankura district. However, the implementation and success of the
policies ultimately depend on the efficiency of the administrators of the local
government, its transparency and willingness to do for the rural women.
6.5. Conclusion
In this chapter we have presented the outlines of the existing policies and their
insufficiencies. Some supplementary and additional policies have been proposed for
further improvement of empowerment level of the rural women and welfare of their
families in the district of Bankura. We conclude this chapter and this study with the
fundamental results of our dissertation as follows.
Women’s empowerment at the household level and at the community level
increases the probability of adopting family planning decision of the rural
people in the district of Bankura.
Household level empowerment of women reduces the probability of
domestic violence against women.
Women’s empowerment at the community level is instrumental to
increasing the share of household income spent for child education.
Age, personal occupation, personal income, financial inclusion, dependency
ratio and highest female education are important determinants of
household level empowerment of the women in Bankura district.
We find age, education of women, access to formal credit, household
landholding, highest female education, the participation in the SHG and
caste as crucial in the determination of women’s empowerment at the
community level.
221
Therefore, women’s empowerments at the household level and at the community level
are instrument of enhancing the household and child welfare of the rural people in the
district of Bankura. However, majority of the sample women do not have admirable level
of empowerment inside and outside home. In order to improve empowerment of women,
which is instrumental for household and child welfare, we have to take some effective
policies that ensure financial inclusion, employment and political participation of the
women and help women undertake income-generating activity. In this regard, micro
enterprise may be the suitable one. Finally, we need to start our journey with fruitful
health and effective education facility towards removing the social and household
practices that act against empowerment and dignity of women. We have to remember the
speech of Swami Vivekananda that a nation which doesn’t respect women will never
become great now and nor will in future. Therefore, to make India a great nation, let us
work towards giving the respect that women deserve in society.
222
BIBLOGRAPHY
Adhikary, M.L. and P. Dutta (2011), ‘Factors Affecting Empowerment of Women: An
Empirical Study in Burdwan, West Bengal’, in Gender Deprivation and Empowerment
of Women An Indian Perspective, Utpal De and Bholanath Ghosh, (eds.) Lambert
Academic Publishing, Germany, pp. 147-163
Adhikary, M.L. and P. Dutta (2012), ‘Mother’s Education and Spending for Children’s
Education: A Micro Study’, Arth Anvesan, Vol. 7(1), pp. 28-38
Adhikary, M.L. and P. Dutta, ‘Women’s Empowerment and Response towards Family
Planning: Empirical Evidence from Bankura District, West Bengal’, Asian Journal of
Research in Business Economics and Management, Forthcoming
Adhikary, M.L. and S. Bagli (2011), ‘Practice of Financial Business through SHGs and
Economic Condition of the Rural Poor – An Empirical Study’, Journal of Practicing
Managers, Vol. 1(2), pp. 08-16
Agarwal, S. (2003), ‘Technology Model for Women’s Empowerment: Reaching the
Unreached’, Kurukhestra, Ministry of Rural Development Vol. 51(7), pp. 18-21
Ahmed, S., M.A. Keonig and R. Stephenson (2006), ‘Effect of Domestic Violence on
Prenatal and Early Childhood Mortality: Evidence from North India’, American Journal
of Public Health, Vol. 96(8), pp 1423-1428. doi 10.2105/AJPH.2005.066316
Alsop, R. and N. Heinsohn (2005), ‘Measuring Empowerment in Practice: Structuring
Analysis and Framing Indicators’, World Bank Policy Research Working Paper 3510,
February 2005,
223
Amin, R., S. Becker and A. Bayes (1998), ‘NGO-Promoted Microcredit Programs and
Women’s Empowerment in Rural Bangladesh: ‘Quantitative and Qualitative Evidence’,
The Journal of developing Areas (Winter 1998), pp. 221-236
Antony, G.M. and K. V. Rao (2007), ‘A Composite Index to Explain Variations in
Poverty, Health, Nutritional Status and Standard of Living: Use of Multivariate
Statistical Methods’, Journal of Royal Institute of Public Health, Vol. 121, pp.578-587
Anuradha., P.S. (2012) ‘An Empirical Study on Socio-Economic Empowerment of
Women through Self-Help Groups’, International Journal of Research in Commerce,
Economics and Management, Vol. 2(1), pp.84-86 Retrieved from www. ijrcm.org.in
Ashraf, N., D. Karlan and W. Yin (2010), ‘Female Empowerment: Impact of a
Commitment savings Product in the Philippines’ World Development, Vol. 38(3),
pp.333-344
Becker, S. (1997), ‘Incorporating Women’s Empowerment in Studies of Reproductive
Health: An Example from Zimbabwe’ Paper presented at the IUSSP seminar on Female
Empowerment and Demographic Processes; Moving Beyond Cairo, Lund, 1997
Beegle, K., E. Frankenberg and D. Thomas (2001), ‘Bargaining power within Couples
and Use of Prenatal and Delivery Care in Indonesia’ Studies in Family Planning, Vol.
32(2), pp.130-146
Bhide, A. (2003), ‘SHGs under Rural Credit Programme: Experience from Maharashtra
Credit: A Key to Prosperity, A Path to Doom’, IASSI Quarterly, Vol. 21(3&4), pp.164-
187
Bhuiyan, M. B. and R. Abdullah (2007), ‘Women Empowerment through
Entrepreneurship Development: Bangladesh Perspective’, Daffodil International
University Journal of Business and Economics, Vol. 2(2), pp. 135-154
224
Bruria, S., R. Hertz-Lazarrowitz and F. Azaiza, (2007), ‘Mothers as Educators: The
Empowerment of Rural Muslim Women in Israel and their Role in Advancing the
Literacy Development of their Children’, International Journal about Parents in
Education, 1 (1), 271-282
Chatterjee, S. (2008), ‘Women’s Empowerment and Role of Self Help Groups- a Socio-
Economic Study in Khejuri (Coastal Bay of Bengal), West Bengal’, Socialist
Perspective, Vol. 36(1-2), pp. 11-29
Chowdhury I. M. and F. Nosheen (2009), ‘The Determinants of Women’s Empowerment
in Southern Punjab (Pakistan): An Empirical Analysis’, European Journal of Social
Sciences, Vol. 10(2), pp. 216-229
Crowell, Daniel W. (2003), ‘The SEWA Movement and Rural Development: The
Banaskantha and Kutch Experience’, SAGE publications India Pvt. Ltd., New Delhi
Dreze, J. and A. K. Sen (1995), ‘India: Economic Development and Social Opportunity’,
Oxford University Press, New Delhi
Dreze, J. and A. K. Sen (2002), ‘India: Development and Participation’, Oxford
University Press, New Delhi
Dyson, T. and M. Moore, (1983), ‘On Kinship Structure, Female Autonomy and
Demographic Behavior in India’, Population Development Review, Vol. 9(1), pp.35-60
Fernando, J. L. (1997), ‘Non-Governmental Organizations, Micro-Credit and
Empowerment of Women’, The ANNALS of the American Academy of Political and
Social Science, Vol. 554(1), pp. 150-177
Ghuman S. J., H. J. Lee and H. L. Smith (2004), ‘Measurement of Women’s Autonomy
According to Women and Their Husbands: Results from Five Asian Countries’, PSC
Research Report, Report No. 04-556, Population Studies Center, University of
Michigan, Retrieved from: http//:www.popupem edu/swaf/papers/rr04.556.pdf.
225
Government of India (1999), Guidelines for Special Projects under SGSY, Retrieved
from: www.rural.nic.in/sgsy.htm
Government of India (2001), Census of India 2001, Ministry of Home Affairs, Office of
the Registrar General and Census Commissioner, New Delhi
Government of India (2001), Introduction to Swayamsidha Scheme, Ministry of Rural
Development
Government of India (2001), Kishori Shakti Yojna, Department of Women & Child
Development, Retrieved from www.wcd.nic.in KSY/ksyintro.htm
Government of India (2001), National Policy for Empowerment of Women, Retrieved
from: mahilakalyan.up.nic.in
Government of India (2011), Census of India 2011, Ministry of Home Affairs, Office of
the Registrar General and Census Commissioner, New Delhi, Retrieved from
www.censusindia.gov.in
Government of India (2011), India Human Development Report, 2011, Towards Social
Inclusion, Institute of Applied Manpower Research, Planning Commission, Oxford
University Press, New Delhi
Government of India (2013), Economic Survey 2011-2012, Department of Economic
Affairs and Economic Division, Ministry of Finance, , Oxford University Press, New
Delhi.
Government of India (2013), Annual Report 2012-13, Ministry of Women and Child
Development
Government of West Bengal (2004), West Bengal Human Development Report,
HDRCC, State Development and Planning Department, West Bengal
226
Government of West Bengal (2005), Statistical Abstract, Bankura District, Bureau of
Applied Economics and Statistics, West Bengal
Government of West Bengal (2007), District Human Development Report, Bankura,
HDRCC, Development and Planning Department, West Bengal
Government of India, NSSO Survey, 1999-2000, & 2009-10, Ministry of Statistics and
Programme Implementation, New Delhi
Gujarati, D. N. (1995), ‘Basic Econometrics’, McGraw-Hill International Editions, New
York
Halvorsen, R. and R Palmquist (1980), ‘The interpretation of Dummy Variables in Semi
logarithmic Equations’, American Economic Review, Vol. 70(3), pp. 474-475
Handy, F. and M. Kassam (2004), ‘Women’s Empowerment in Rural india’, Paper
Presented on the ISTR conference, Toronto, Canada, Retrieved from:
http://www.jhu.edu/~istr/conferences/toronto/workingpapers/handy.femida.pdf
Hossain, I., T. Akhtar, M. M. Rahman, and M. T. Uddin, (2006), ‘Rethinking the
Approaches to Women Empowerment: Bangladesh Perspective’, Journal of Applied
Sciences, Vol. 6(5), pp. 1159- 1164
International Institute for Population Sciences (IIPS) and Macro International (2007),
India National Family Health Survey (NFHS-3), 2005–06: Volume I. Mumbai: IIPS
Jain, R., R.K. Khushawa, and A.K. Srivastav (2003), ‘Socio-Economic Impact through
Self-Help Group’, Yojana, Vol. 47(1), pp. 11-12.
Janssens, W. (2010), ‘Women’s Empowerment and the Creation of Social Capital in
Indian Villages’, World Development, Vol.38 (7), pp 974-988
227
Jejeebhoy, S. J. (1995), ‘Women’s Education, Autonomy, and Reproductive Behaviour:
Experience from Four Developing Countries’, International Studies in Demography,
IUSSP, Oxford: Clarendon Press
Jejeebhoy, S. 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, Harriet Presser and Gita Sen, (eds.)
New York, Oxford University Press.
Jejeebhoy, S. J. and Z. A. Sathar (2001), ‘Women’s Autonomy in India and Pakistan:
The Influence of Religion and Region’ Population and Development Review, Vol. 27
(4), pp. 687-712
Johnston, J. (1984), ‘Econometric Methods’, 3rd Edition’, McGraw-Hill Book Company,
New Delhi
Kabeer, N. (1999), ‘The Conditions and Consequences of Choice, Reflections on the
Measurement of Women’s Empowerment’, Discussion paper, DP 108, United Nation
Research Institute for Social Development (UNRISD), Geneva.
Kabeer, N. (2001), ‘Resource, Agency, Achievement: Reflections on the Measurement
of Women’s Empowerment” in ‘Discussing Women’s Empowerment – Theory and
Practice’ Edited by Anne Sisask, SIDA Studies No. 3, Novum Grafiska, AB, Stockholm,
pp. 17-57.
Kaladhar, K. (1997), ‘Microfinance in India Design, Structure and Governance’
Economic and Political Weekly, Vol. 32(42), pp. 2687-2706.
Khan, M. R. and F. Ara (2006), ‘Women, Participation and Empowerment in Local
Government: Bangladesh Union Parishad Perspective’, Asian Affairs, Vol. 29 (1), pp 73-
93.
228
Kishor, S. (1997), ‘Empowerment of Women in Egypt and Links to the Survival and
Health of their Infants’, Paper presented at the seminar on female empowerment and
demographic Processes, International Union for the Scientific Study in Population.
Kishor, S. (2000), ‘Empowerment of Women in Egypt and Links to the Survival and
Health of their Infants’, in Women’s Empowerment and Demographic Processes:
Moving Beyond Cairo, Harriet Presser and Gita Sen, (eds.), Oxford University Press,
New York.
Koenig, M.A., S., Ahmed, M.B. Hossain and A.B.M.K.A. Mozumder (2003), ‘Women's
Status and Domestic Violence in Rural Bangladesh: Individual- and Community-Level
Effects’, Demography, Vol., 40(2), pp. 269-288
Koenig, M. A., R. Stephenson, S. Ahmed, S. J. Jejeebhoy and J. Campbell (2006), ‘
Individual and Contextual Determinants of Domestic Violence in North India’, American
Journal of Public Health, Vol 96(1), pp. 132-138.
Kumar, N.P. (2011), ‘Status of Women and Economic Development A Decadal Analysis
based on Indian States’ Socio-economic Voices, Retrieved from www.
indiasta.com/article/35nomita/fulltext.pdf
Lillykutty, Sr. (2003), ‘Education and Empowerment of Women Enhance Quality of
Life’, IASSI Quarterly, Vol. 21(3 & 4), pp. 234-243
Maldonado, J.H., C. Gonzaler-Vega and V. Romero, (2003), ‘The Influence of
Microfinance on the Education Decision of Rural Households: Evidence from Bolivia’,
Paper presented at the Annual Meeting of the American Agricultural Economic
Association, Montreal, Canada 2003, pp. 27-30
Mason, K.O. and H.L. Smith, (2003), ‘Women’s Empowerment and Social Context:
Results from Five Asian Countries’, Gender and Development Group, World Bank,
Washington, D.C.
229
Mayoux Linda (2000), ‘Microfinance and the Empowerment of Women: A Review of
the Key Issues’, Social Finance Working Paper No. 23, International Labour
Organization, Geneva, Retrieved from
J:\ilo_data\public\english\employment\finance\download\wp23.wpd
Meher, S.(2007), ‘Impact of Micro-Finance on Poverty A study of Self-Help Groups in
Orissa’, Journal of Rural Development, NIRD, Hyderabad, Vol. 26 (3), pp. 315-333,
Mehra., R (1997), ‘THE ROLE OF NGOs: CHARITY AND EMPOWERMENT:
Women, Empowerment, and Economic Development’, The Annals of The American
Academy of Political and Social Science, Vol. 554, pp. 136-193
Molhotra, A. and M. Mather (1997), ‘Do Schooling and Work Empower Women in
Developing Countries? Gender and Domestic Decision in Sri Lanka’, Sociological
Forum, Vol. 12(4), PP. 559-630
Molhotra, A., R.S. Sidney and C.B. Boender, (2002), ‘Measuring Women’s
Empowerment as A Variable in International Development’ the Gender and
Development Group of the World Bank and the World Bank’s Social Development
Group, pp. 1-47, Retrieved From: www.aed.org/LeadershipandDemocracy/
upload/MeasuringWomen.pdf
Moser, C. and A. Moser (2005), ‘Gender mainstreaming since Beijing: a review of
success and limitations in international institutions’, Gender and Development, Vol.
13(2), pp. 11-21.
Murthi, M., A-C Guie and, J. Dreze (1998), ‘ Mortality, Fertility and Gender Bias in
India’ in INDIAN DEVELOPMENT Selected Regional Perspectives, J. Dreze and A. K.
Sen (eds.), Oxford University Press, New Delhi, pp. 357-399.
Mustafa, G., I. Tareque, M. Haque and T. M. Islam, (2008), ‘Mathematical Modeling of
Women’s Empowerment in Bangladesh’, Research Journal of Applied Sciences, Vol.
(6), pp. 416-420
230
National Commission for Women, India (2014), website: ncw.nic.in
Narasimham, S. (1999): ‘Empowering Women: An Alternative Strategy for Rural India’,
Sage publication, New Delhi.
National Crime Records Bureau (2011 & 2013), ‘Crime in India Statistics’, Ministry of
Home Affairs, Government of India. New Delhi, Retrieved from: http://ncrb.gov.in
Nayak, P. and B. Mahanta (2009), ‘Women Empowerment in India’, MPRA Paper No.
12685, Retrieved from: http://mpra.ub.uni-muenchen.de/12685
Olumakaiye, M.F. and A.O Ajayi (2006), ‘Women’s Empowerment for Household Food
Security the Place of Education’, Journal of Human Ecology, Vol. 19(1), pp. 51-55
Panda, B. and B.K. Pattnaik, (2003), ‘Role of Grassroots NGO: A Social Activities
Perspective’, Journal of Rural Development, NIRD, Hyderabad, Vol. 22 (1), pp. 21-44
Parveen, S. and I.U. Leonhauser, (2004), ‘Empowerment of Rural Women in
Bangladesh: A Household Level Analysis’, Conference Paper, Conference on Rural
Poverty Reduction through Research for development and transformation, Deutscher
Tropentag-Berlin, 5-7 October 2004, pp.1-10. Retrieved from: http//www.tropenlog
de/2004/abstract/full/382/pdf
Parvin, G.A., S.R. Ahsan and M.R. Chowdhury (2004), ‘Women Empowerment
Performance of Income Generating Activities Supported by Rural Women Employment
Creation Project (RWECP): A Case Study in Dumuria Thana, Bangladesh’, The Journal
of Geo-Environment, Vol. 4, pp. 47-62.
Pitt, M. M. and S.R. Khandker (1996), ‘Household and Intra-household Impact of the
Grameen Bank and Similar Targeted Credit programs in Bangladesh’ World Bank
Discussion Paper No. 320
231
Pitt, M. M. and Khandker, S. R. (1998), ‘The Impact of Group-Based Credit Programs
on Poor Households in Bangladesh: Does the Gender of Participants Matter?’, Journal
of Political Economy, Vol. 106(5), pp. 958-96
Planning Commission, India (2010), Statistics on Women in India 2010, Five Year Plans
New Delhi
Planning Commission (2011), Faster, Sustainable and More Inclusive Growth An
Approach to the Twelfth Five Year Plan (2012-17), Government of India. New Delhi,
Retrieved from: www.planningcommission.nic.in
Quisumbing, A. R. and J. A. Maluccio (1999), ‘Intrahousehold Allocating and Gender
Relations: New Empirical Evidence’ Policy Research Report on Gender and
Development’, Working Paper Series No.2, World Bank Development Research
Group/Poverty Reduction and Economic Management Network, Washington
Quisumbing, A. R. and Benedicte de la Briere (2000), ‘Women’s Assets and
Intrahousehold Allocation in Rural Bangladesh: Testing Measures of Bargaining Power’,
Food Consumption and Nutrition Division Discussion Paper No. 86, International Food
Policy Research Institute, Washington.
Ray, K., M. Chakraborty, H. Roy, S. Gupta and I. Banerjee (2012), ‘Violence against
Women’: Evidence from a Cross Sectional Study in Urban Area of North Bengal’, Al
Ameen Journal of Medical Science, Vol. 5 (2), pp 157 -64
Reddy, G.N. (2002), ‘Empowering Women through Self-Help Groups and Microcredit:
The Case Study of NIRD Action Research Project’, Journal of Rural Development,
NIRD, Hyderabad, Vol. 21 (4), pp. 511-535.
Rocca, Corinne H., S. Rathod, T. Falle, R. P. Pande and S. Krishnan (2008),
‘Challenging Assumptions about Women’s Empowerment: Social and Economic
Resources and Domestic Violence among Young Married Women in Urban South
India’, International Journal of Epidemiology, Vol.38, pp. 577-585.
232
Rowlands, J. (1995), ‘Empowerment Examined’, Development in Practice, Vol. 5(2),
pp. 101-107.
Samanta, G. (2009), ‘Microfinance and Women: Gender Issues of Poverty Alleviation
and Empowerment’, The MicroFINANCE REVIEW, Vol. 1(1), pp. 100-120.
Santillán, D., R. S.Sidney, T. A. Hoang; H. M.Tran, T. T. Quach and M. D. Nguyen
(2004), ‘Developing Indicators to Assess Women’s Empowerment in Vietnam’,
Development in Practice, Vol. 14(4), pp. 534-549
Sarker, M. (2010), ‘A Study on Domestic Violence against Adult and Adolescent
Females in a Rural Area of West Bengal’, Indian Journal of Community Medicine,
35(2), pp. 311–315
Sen, A, (1998), ‘Social Exclusion an Alternative Concept for the Study of Deprivation’,
IDS Bulletin, No.1
Sen, A, (1999), ‘Development as Freedom’, Oxford University Press, Oxford
Singha Roy, D. K. (1995), ‘Peasant Movement and Empowerment of Rural Women’,
Economic and Political Weekly, Vol. 30(37), pp. 2306-2311
Sridevi, T.O. (2005), ‘Empowerment of Women: A Systematic Analysis’, IDF
Discussion Paper, Retrieved from: http//www.idfresearch.org /pdf /0411.p.d.f.
Swaminathan, M. (1990), ‘Village Level Implementation of IRDP Comparison of West
Bengal and Tamil Nadu’, Economic and Political Weekly, Vol. 25(13), pp. A17-A28
Tripathi, T., (2011), ‘Women’s Empowerment: Concept and Empirical Evidence from
India’, Paper presented in a National Seminar, Delhi School of Economics, Retrieved
from: www.ededse.org/ws2011/papers/Tulika%20 Tripathti.pdf.
233
UCLA (2010), ‘Annotated SPSS Output Principal Components Analysis’ Academic
Technology Services, Statistical Consulting Group, Retrieved from
www.ats.ucla.edu/stat/SPSS/output/principal_components.htm
UNICEF (1993), ‘Central and Eastern Europe in Transition: Public Policy and Social
Conditions’, Regional Monitoring Report No. 1, UNICEF, New York
United Nations Development Programme (1990), ‘Human Development Report 1990’,
Oxford University Press, New York.
United Nations Development Programme (1995), ‘Human Development Report 1995’,
Oxford University Press, New York
United Nations Development Programme (2003-2011), ‘Human Development Reports
2003-2011’, Oxford University Press, New Delhi
United Nations (1993), ‘A/RES/48/104 Declaration on the elimination of Violence
against Women, United Nations General Assembly’, Retrieved from
www.un.org/documents/ga/res/48/a48r104.htm
United Nations (2003), Indicators for Monitoring the Millennium Development Goals,
ST/ESA/STAT/SER.F/95, United Nations Publication, Sales No. E.03.XVII. 18,Retrieved
from: mdgs.un.org/unsd/mdg/Resources/Attach/Indicators/HandbookEnglish.pdf
United Nations (2008), The Millennium Development Goals Report 2008, New York:
United Nations Retrieved from:
http://www.un.org/millenniumgoals/2008highlevel/pdf/newsroom/mdg%20reports/MDG
_Report_2008_ENGLISH.pdf
Varadarajan, D.B. & K.A. Raj (2002), ‘Achieving Women’s Empowerment through
Women’s Entrepreneurship’, IASSI Quarterly, Vol. 21 (1), pp. 97-111
234
Varghese T (2011), ‘A Study based on Women Empowerment Index’, Far East Journal
of Psychology and Business, Vol. 2(2), pp. 37-53, Retrieved from:
www.faeastjournals.com/files/V2N2P3.pdf
Vasimalai, M. and Narender, P. K. (2007), ‘Micro-Finance for Poverty Reduction: The
Kalanjiam Way’, Economic and Political Weekly, Vol. 42(13), pp.1190-1195
Verma, M. M. (2009), ‘Approaches to Women’s Empowerment: An Overview’ IASSI
Quarterly, Special Issue, pp. 230-249.
Viswanathan, R. (2001), ‘Development, Empowerment and Domestic Violence
Karnataka Experience’, Economic and Political Weekly, Vol.36 (24), pp. 2173-2177.
Williams, J. (2005), ‘Measuring Gender and Women’s empowerment Using
Confirmatory Factor Analysis’, Working Paper, Research Program on Population
Processes, Institute of Behavioral Science, University of Colorado, Boulder, Retrieved
from: www.colorado.edu/ibs/pubs/pop/pop2005-0001.pdf
World Bank (2001), ‘World Development Report, 2000/2001, Attacking Poverty’,
Oxford University Press, New York
World Bank (2001), ‘World Development Indicators’, World Bank Publications,
Retrieved from: www.scribd.com/doc/16060137/World-Development-Indicators-2001
World Bank (2012), ‘World Development Report, 2012, Gender Equality and
Development’, World Bank, Washington, DC
Yunus, M. and A. Jolis, (1998), ‘Banker to The Poor: The Autobiography of Muhammad
Yunus, the founder of the Grameen Bank’ The University Press Limited, Bangladesh
Zaman, U.R., M.M. Rahaman, S.M.A. Hussain and M. Zaki (2012), ‘Women
Empowerment in Different Household Issues of Bangladesh’ Bangladesh Medical
Journal, Vol. 37(2), pp. 41-45