In his ‘re-examination’ of inequality, Sen (1992) argues that the space of outcomes is a more appropriate space for examining inequalities rather than the space of commodities or entitlements.
These outcomes could be social, economic or political, and it is possible to identify suitable indicators in these to discover how a group—women in our case—fares compared to the other population at large. The process of governance has an impact on all these outcomes that together define the well-being of its citizens.
In judging the well-being of women, the different components of well-being need to be examined. One can begin with a very basic outcome, i.e. survival itself, reflected in life expectancy and in the sex ratio at birth. This can be followed by the quality of survival; whether life is ‘short, nasty and brutish’, or otherwise. Then comes the question of the skills of social reproduction one is able to acquire and utilise. Is one able to educate oneself? Is the productive use of a skill impaired? Is the skill remunerated equitably? The issue of participation in the social sphere comes next. Is one able to participate in the social sphere? Is the participation equitable? Where does one stand in the power structure in respect of such participation? The list can be expanded further.
It is necessary and useful to ask how men and women fare in respect of each of these outcomes. How does their longevity compare? Are they equally well-nourished? Are their skill attainments comparable? Is there a disparity in their wages? How are they represented in different domains of public life?
If the social processes were equitable, the ‘distribution’ of men and women in terms of most of these parameters would be similar. But, in reality, social processes are not quite equitable and this reflects in different aspects of well-being: the span of life, its quality, access to resources, among other factors. If we draw, at random, a sample of men and women and study any of these characteristics, we will find men and women clustered separately, revealing a gender inequality gap in respect of that characteristic.
But, in reality, social processes are not quite equitable and this reflects in different aspects of well-being: the span of life, its quality, access to resources, among other factors. If we draw, at random, a sample of men and women and study any of these characteristics, we will find men and women clustered separately, revealing a gender inequality gap in respect of that characteristic.
These gender inequality gaps may also be sample-specific. Gender is not the only identity a person has. There are many others: regional, religious, caste or class, to name a few. Among these groups, the gender inequality gaps revealed by different indicators of well-being may vary significantly. The gender inequality gap in nutrition may differ by region; sex ratios may differ by class; gaps in educational levels may differ by religion. We need, therefore, two separate levels at which the data are disaggregated by gender: one for different aspects of well-being and the other for different relevant groups to which men and women belong.
The search for useful indicators to assess gender sensitiveness in governance can either precede the efforts to bring greater awareness about gender equity or follow these. In one case, a given indicator, e.g. the sex ratio in a region, can reveal the inequality of survival while in the other, an indicator such as the proportion of women represented in elected bodies can be used for monitoring how society translates its rhetoric into practice.
Finally, gender-based disaggregation of development indicators is necessary to distinguish between ‘development’ and ‘discrimination’. A gender-blind governance may only focus on the overall indicator, e.g. under-five mortality rate, as a measure of development, but overlook the discrimination hidden within it, i.e. the gender inequality gap. A society may have high levels of both development and discrimination, but that is not the desirable goal. Society has to strive for high levels of development combined with a low level of discrimination. The use of gender-disaggregated indicators allows us to monitor if this is happening.
Different aspects of well-being
Our concern for gender equity need not stray into the domain of definitions—what constitutes governance? or how perfect is a given definition of well-being? There are accepted indicators of well-being
that are presently used. These should be scrutinised first for gender equity. To begin with, there is no harm in assuming that whatever is good for men, is good for women as well. A ‘fine-tuning’ of the indicators of well-being can follow later.
Mortality rates are a good indicator for levels of development. All other things being equal, lower mortality would usually correspond to higher development. The following indicators can be used for monitoring different aspects of mortality among men and women:
- Crude death rate
- Under-five mortality rates
- Infant mortality rates
- Still-birth rates
- Maternal mortality rates
- Sex ratio at birth
These indicators assume importance in the Indian context, given the high masculinity of its sex ratios and the increasing ‘female deficit’. Further, this deficit arises largely on account of excess female mortality under the age of 5 years. Female children are under threat from sex-selective elimination at the foetal stage, infanticide at birth, excess mortality in infancy and in the 1-4 years’ age group. Each of these disadvantages is behavioural in nature and not biological.
Increasingly, the masculine sex ratio at birth (SRB) indicates a higher incidence of sex-selective abortion. The extent of excess female mortality in the 1-4 years’ age group indicates the extent of gender bias against girl children. Excess female infant mortality indicates its intensification.
Maternal mortality rates (MMR) are an indicator of the extent to which a society cares for mothers. The rates of MMR in India are unusually high and, more important, most of these are avoidable.
Quality of survival
Among the women who survive, life expectancy, access to different health facilities and nutritional status become important concerns so that life is less ‘short, nasty and brutish’. In this context the following indicators are relevant:
- Life expectancy at birth
- Immunisation coverage
- Nutritional status
- Age at marriage
- Age at first pregnancy
While life expectancy at birth is a much analysed and monitored indicator, the same cannot be said in respect of immunisation and nutritional status. The asymmetry in nutritional status is either considered to be ‘obvious’ or not worth bothering about. The age of a woman at marriage/first pregnancy has a bearing on her health status and on the incidence of low birth weight. It is necessary to monitor this indicator closely to locate the emergence of child marriage. It may also become necessary to look at the distribution of age at marriage in different regions/ communities to gain a better understanding of the social dynamics of the process.
The next important aspect of well-being relates to acquiring skills for social reproduction. Achievements in literacy and levels of education are appropriate indicators for these. However, there could be a considerable gender inequality gap in these levels, starting with:
- Dropout at primary level
- Completion of primary education
- Dropout at secondary level
- Completion of secondary and higher education.
Inequality in this sphere is one area where a specific focus by ethnic groups is also needed. Further, women’s groups also need to take up female literacy campaigns in small pockets aiming at saturation. Female education up to the primary level has so many spin-offs that a separate focus outside government efforts is called for.
The issue of dropouts needs to be studied at different stages of education. It is quite likely that the tendency to drop out is strong up to a stage and beyond it the level diminishes. Such information is extremely useful for policy.
Acquisition of skills may still not ensure engagement with the labour market or equitable terms of engagement. Women’s workforce participation is a contentious area, and needs serious monitoring. The relevant parameters include:
- Workforce participation ratios
- Patterns of workforce participation
- Wage disparity
- Waged and unwaged work
- Workplace conditions, including safety at workplace
- Patterns of migration
- Women-headed households
The monitoring of these indicators assumes importance given the close correlation between survival and workforce participation. Papanek (1989) has written extensively about the process of withdrawal of women from waged workforce participation for the purpose of ‘status production’ for the family. Such withdrawal makes it difficult for the woman concerned to get back to work if she needs to. This is particularly so for skilled and white collar jobs. Further, the prosperity of a household may adversely affect the status of female members if their contribution to such prosperity is nil or not recognised (Agnihotri, 2000).
Studying the migration patterns and changing nature and composition of female-headed households assumes importance here. This is especially so in the context of globalisation of the economy and ‘feminisation’ of poverty.
Control over resources
While high female workforce participation rates definitely improve the status of women, control over resources is a different story altogether. In this context, one needs to study the following:
- Participation in surplus-generating activity
- Control over resources/means of production
- Land ownership
- Property rights
It seems that women are usually allowed to engage in subsistence-level income generation. However, their ability to move to the surplus domain and exercise control over means of production, whether capital or land, gets severely restricted (Ram, 1991; Agarwal, 1994).
Participation in the public sphere
Basu (1992) has given a very simple but effective description of the determinants of the status of women. She looks at three factors: exposure to the ‘external sphere’, participation in it, and the decision-making ability during the participation. It is difficult to evolve indicators for these factors. But to begin with, the sex ratio of voters vis à vis the sex ratio in the population above 18 years of age can be one useful measure of exposure, while the sex ratio in the votes cast can be a useful measure of participation. The stability of the tenure of the elected representatives in local bodies can indicate the degree of freedom in decision-making.
It is not just the ballot-based political process that represents the ‘public’ sphere. Women’s participation in extra-familial groups such as self-help groups (SHGs), trade unions and even political parties needs to be studied in detail.
Women’s participation in the public sphere considerably depends upon the security or the law and order environment. Whether a college-going girl feels free to travel by bicycle in the city will depend, to a large extent, on her own perception of safety. It is, therefore, necessary to analyse the incidence of crime against women compared to the general level of crime. It would bear reiteration that a rise in crime will affect the vulnerable sections more, and it should not be surprising that the adverse effect on women will be greater. District-level crime data available from the National Crime Research Bureau (NCRB) will be a useful source in this regard. It is necessary to monitor, in particular, the following:
- Overall incidence of crime against women
- Rape and molestation
- Dowry and marriage-related violence and murder
- Unnatural cases of death below and above 20 years of age
It has been seen that the general crime pattern and the pattern of crime against women differ significantly. Further, different crimes, e.g. dowry-related murders and those related to sexual offences, show different trends. Certain regions contribute disproportionately to these crimes as compared to their share in the population.
Some methodological issues
A question may arise as to what new insights, if any, can such an analysis bring. There are two issues here: one procedural and the other methodological. The procedural issue relates to cataloguing relevant indicators and ensuring that gender disaggregated data on this are collected and made available. The extent to which data collection can be gender-blind can surprise many! Many prestigious ‘surveys’ are a case in point. One often comes across separate figures by sex—male or female; location—urban or rural; and caste—SC or ST. But more often than not, within the urban or rural break-up, or within the SC or ST category, a break-up by gender is not provided. Is it that the differences are not significant? Hardly anyone will believe that. Yet this omission is apparent in survey after survey.
All these indicators are illustrative, but they capture major aspects of well-being. There is definitely scope and a need to identify further indicators which can fine-tune the ‘search’. But that task can take place separately.
The second issue is methodological. Making the data available often becomes an end by itself. Its analysis either does not follow, or follows in a leisurely fashion; and when it does, it is not policy-oriented but oriented to the academic perfection of the ‘second decimal point’. However, thanks to developments in IT, many techniques are available today which can provide useful insights at the policy level through the analysis of the available data.
Geographical mapping is one such technique. It can identify regional clusters where particular social patterns are strong or weak (Agnihotri, 1996; 2000). It can identify, for example, contiguous tracts of high girl-child mortality or high female literacy for that matter. Such ‘contour mapping’ of social variables can have strong synergy with policy formulation exercises.
Another useful technique is to look at the way in which the values of any relevant variable differ by gender compare, and then its deviation from the hypothetical situation where men and women were ‘identical’ in nature. This technique can be used for longitudinal as well as cross-sectional data. A study of the nature of deviation mentioned above can provide useful insights about the gender bias (Agnihotri, 1999; 2001).
Together, these techniques can reveal the compounding disadvantages among certain groups, e.g. adverse survival chances of scheduled caste girl children in north-western states, or the unusually high male infant mortality in parts of the central tribal belt in the country (Agnihotri, 2000).
Gender-sensitive versus gender-blind governance
The need to develop these indicators is particularly strong in our society where the gender bias against girl children and younger women is well known and adequately documented. India’s de jure track record about gender equity in governance is quite good. The problem arises in translating this into reality. Where governance is good in preaching gender equity but indifferent about practising it, relentless monitoring of relevant indicators is a must. All the more so when the apparatus of governance is ‘manned’ by well meaning but gender-blind persons — women included.
Those who might find this description hypothetical, uncharitable or harsh, may consider the following:
- An exhaustive survey of certain artisans provides figures
about the family composition of these artisan households
in terms of ‘adult males’, adult females and children. This
makes it impossible to calculate the sex ratio, either for the
total population or for the child population (useful since it
is free of sex-selective migration). The statistician in charge
of designing the survey format was a woman. Her remark
was that this problem ‘never occurred to her’.
- An exhaustive all-India survey on the health of children
in a primary school never provided the data by gender.
The response of state government officials was that the
proforma prescribed by the central government did not
‘prescribe’ it. The ailments surveyed included anaemia, a
problem known to affect girls more than it does boys!
- A well-meaning senior officer had argued that providing
a male-female break-up in terms of the nutritional status
of children in the Integrated Child Development Services.
(ICDS) programme involved ‘too much trouble’, and that this break up should be dispensed with to make reporting simpler! When this actually happened, there was hardly any protest; academicians did not bother, activists did not even come to know about it.
Does one have to repeat ad nauseum that eternal vigilance i the price for liberty (in this instance even survival)?