On Measures of Empowerment: The Gender Aspect
Rabindranath Mukhopadhyay[*] and Jaydip Datta[†]
Abstracts
Two major varieties of human empowerment could be thought of: (a) economic empowerment & (b) social empowerment. Higher is the position in one’s career ladder, higher is one’s economic empowerment. Social empowerment reflects one’s decision making capacity and capacity to appreciate merits and demerits of decisions arrived at by others. Both the varieties of empowerment are important to both men and women. A multidimensional measure of empowerment, as an extension of Watt’s measure, following (Chakravarty, et. al., 2005), has been constructed explicitly incorporating this gender dimension. Further, a generalization of this measure, accommodating n-dimensions, has been demonstrated. Moreover, utilising additive seperability property, the said measure has established how between-group divergence in women empowerment could be captured. This has made the present measure more useful for the policy exercises. It has been possible to demonstrate, with the help of Indian data, how changes in the value of measure, over time or over space, are relatively influenced by values of two sub-measures corresponding to the said two varieties of empowerment. Empirical illustration supporting additive decomposability property of the present measure could not be accommodated due to lack of suitable data.
1. INTRODUCTION: DevelopmentEconomicshas always been concerned to find out the path of development for the underprivileged nations so that they can uplift their economy at par with the developed nations of the world. However, since the introduction of the annual series of the Human Development Report (UNDP, 1990), for every member nation, quality of life has become equally important with the macro economic growth. Accordingly, empowerment of people has become an important variable of concern (Mason, 2005; Anand & Sen, 2003). By empowerment one broadly refers to the expansion of freedom of choice and action to transform one’s current life into a preferred one. It does not require to explain that concern for improvement is more crucial for the poor or developing nations than with the other part of the world. Again, the problem becomes more acute if one takes into account explicitly the gender aspect. In the developing world, lack of empowerment of women is inhibiting the desired progress in the quality of life. Hence improvement in empowerment of women has to be emphasized to have better quality of life.
In the contemporary world, women and men share many aspects of living together, collaborate with each other in complex and ubiquitous ways, both within and outside home, and yet end up with different rewards and deprivations. As a result, there is a strong case for a more comprehensive investigation of gender inequity in economic & social developments of the poor nations (Anand & Sen,2003).
Such as an exercise will become more meaningful if it can successfully identify efforts & contributions made by women that go unrecognized in standard national income & employment statistics.
This paper is an attempt towards measurement of empowerment from the gender perspective so that backwardness of women in the development process gets adequately explained.
2. THE PROBLEM:To overcome the gender insensitivity of the Human Development Index (HDI), UNDP has introduced gender related development index (GDI) along with HDI. It has also introduced simultaneously Gender Empowerment Measure (GEM), which provides a measure of gender inequality in the areas of agency and power. Both are useful for policy analysis and analytical research.
The GDI does not measure the extent of gender inequality in itself. Thus, there exists a need to realize gender inequality that accompanies absolute levels of human development (Dijkstra & Hanmer, 2000). Whatever, be the absolute level of human development, a high degree of gender inequality is an ethical problem and should concern government. Furthermore, understanding of the relationship between gender equality and general welfare can only be advanced if a measure of gender inequality is available. Thus GDI can be considered as a step towards capturing gender–related development.
Sustainable reduction in gender inequality is possible only when there is increase in women empowerment. This very fact has been highlighted in the works of many economists (Anand & Sen, 2003; Mason, 2005 ). Economists like Mason, opines that empowerment is basically a property of social & cultural systems rather than that of individual experiences and traits (Smith, 1989). It is further argued that empowerment is multidimensional (Mason, 2005.). Here the multidimensional character of women empowerment is hidden in various dimensions of empowerment such as financial empowerment, cultural empowerment etc. Financial empowerment usually provides women the much sought after economic independence thus increasing their freedom of action with regard to economic affairs of the family. On the other hand, cultural empowerment provides women with the power to have a better understanding of the world environment prevailing both within and outside the household. This guides them to acquire the skill necessary to increase their productive efficiency. This also enables them to structure their family and provide attention to their wards, especially girl children, so that the future generation can lead a better quality of life, at least with respect to gender discrimination. All the above calls for developing a multidimensional empowerment measure with a gender aspect embedded into it in order to generate an insight into the problem of gender inequality.
In the present paper an attempt has been made to provide such a measure along with empirical illustrations. Section 3 provides an outline of the motivation and the approach followed for the construction of such a multidimensional women empowerment measure. The actual construction of the measure is described in Section 4. Utility of such a measure with the help of empirical exercise has been discussed in Section 5. Finally, conclusion along with limitations has been stated in Section 6.
3. MOTIVATION & THE APPROACH FOLLOWED:Following Sen (1976), any measure of poverty is primarily concerned with the extent of poverty instead of counting the number of poor. The poverty index provided by Watt (Tsui 2002) has accommodated multidimensional character of poverty. Further, it has been established that Watt’s poverty measure is one of the most well-behaved multidimensional measure (Chakraborty, et. al. 2005).
Since empowerment is a multidimensional variable, under the influence of the said development of literature, a measure of empowerment, incorporating the gender aspect, has been attempted here. To make the measure self contained, a standard neo-classical general equilibrium framework has been utilized. The proposed measure has been demonstrated as a close variant of Watt’s measure of poverty. In this measure the weights are endogenously determined. The economic significance of these weights will be discussed in due course.
A novelty of the proposed measure is that it is capable of capturing changes over time in women empowerment. As a result, it is useful for purposes of policy planning. The inter-linkage between different components of the proposed measure and changes in the said linkage over time could be utilized for policy exercises. This is another merit of the proposed measure.
4. THE MEASURE DEVELOPED:Empowerment being multidimensional, calls for interdependence among its various dimensions. Further, there exists interdependence within each sub-sector of these dimensions, too. Hence, construction of the proposed measure is carried out in a standard neo-classical general equilibrium framework. Two broad categories of empowerment - (i) financial empowerment, and (ii) cultural empowerment - are well recognised in the literature. [Mason, 2005; Sen & Anand 2003]
Financial Empowerment: It is widely accepted that one of the major ways by which women can financially empower themselves is to get engaged in employment (Mason, 2005). Women having income of their own, are believed to be more empowered than others. As observed in Bangladesh micro-credit program, husbands resort to violence when their wives borrow and setup independent economic enterprises, perhaps because, these enterprises allow these women to become more independent from and less submissive to their husbands (Rahman 1999). Thus employment of women can very well be used as an indicator of female empowerment. Again within employed women, further categorization is possible according to income classification – higher the income earned, higher is the financial empowerment. There is no denying that income earned, varies with skill. Without any loss of generality, it may be assumed that, activities in general, (barring a few exceptions, like managerial services) services in primary sector are less skill demanding than their counterparts in secondary and tertiary sectors. Further, there is seasonal variation in employment in primary sector.
0This feature is true only for the unorganized sectors, as far as employment in secondary and tertiary sectors are concerned.
Additional difficulty of employment of women in cultivation is that barring routine operations, all other activities, are not physically suitable for them. 1As a result, engagement of women in the former (barring the case of specialised jobs) has a higher probability of getting associated with low income, compared to it’s counterpart in the latter two sectors. Thus higher the number of women employed in the manufacturing and services sectors, higher is the financial empowerment of women in the society concerned. Hence by examining the distribution of employment of women across the above-mentioned three-macro sectors, it is possible to infer about the nature of financial empowerment of women in the concerned society.
These income differentials between sectors are likely to have demonstration effects – both short run as also long run. Those who have already acquired skills would make attempts to get them better placed; i.e., they would try to migrate out from low paid sector to high paid sectors, as far as possible. 2 This feature could be identified as the short run effect. On the other hand, those who are lacking the requisite skills, necessary for such an income benefiting migration between sectors, would take time to upgrade their skills and to encash it accordingly. As a result, this phenomenon could be called as the long run effect. To accommodate both these effects adequately, it is better to consider the unit of measurement for time as a decade. Employment figure for any decade is nothing but the difference between employment figures of the terminal year and the base year of the decade concerned.
1 Such a possibility of restriction on employment of women within secondary sector, for example, in engineering industries, is a reality.
2Again, such a migration is even possible between sub-sectors within any sector.
When employment figures separately for these three sectors for any decade are compared, the above-mentioned short run and long run effects are adequately accommodated. Here lies the benefit of accepting decade, instead of any lower denomination like year, month, week, day, etc., as the unit of measurement of time for the current purpose. Time series analysis of sectoral employment figures for a few consecutive decades will speak about changes over generations.
Let Ijt be the percentage of females in the total workforce engaged in the jth sector at the tth period; j=1,2,3 and t=1,2,…..,T. Further let j=1 represent the primary sector and similarly j=2 and 3 represent secondary and tertiary sectors respectively.
Functional representations of the above mentioned features, in a general equilibrium framework, could be assumed to be of the following Cobb-Douglas varieties:
:I1t = I2t2t I3t3t …(1)
I2t = I3t3t I1t1t …(2)
I3t = I1t1t I2t2t …(3)
It is thus a determinate system. From a given set of values of I1t, I2t & I3t corresponding to time period ‘t’, the values of the parameters 1t, 2t3t are obtained by solving the above equations (1), (2) & (3).Here, αjt represents importance or weight assigned to the jth sector for the tth period. Dispersion of women engagement across these sectors will determine the weight to be assigned to the concerned sector. Thus the sector having lower percentage of women engagement will get higher weight. At the early stages of development, activities in the primary sector will be higher than those in the secondary sector, which in turn will be higher than those in the tertiary sector. As a result, at the early stages of development, in terms of contribution to GDP, primary sector has a bigger role followed by secondary and tertiary sectors in this order (Rostow,1960). Thus more women are likely to be engaged in the primary sector followed by secondary sector and tertiary sector in this order. This distribution of employment of women will be the representative picture for any poor nation at the early stages of development. Accordingly, the relationship between the weights will be of the order 1t2t3t. This fact is provided by the very nature of interdependence of I1t, I2t & I3t. However, with the progress of development, order of the sector specific contributions to the GDP will change: relative contribution from the primary sector will decline and correspondingly, relative contributions of the other two sectors will increase. At the later stages of development, contribution to GDP will be maximum from the tertiary sector, followed by the secondary sector and least from the primary sector. Under its influence the dispersion of employment of women across all the sectors will get reduced. This in turn, will lead to equalization of weights across sectors. Thus, the weights here get adjusted according to the relative importance of the Ijt . It may be noted here that higher weight is given on lower value of the indices and vice versa. The higher weight highlights the attention required to the poor performance of the indices. Ideal situation will be represented by the case when all the sector specific indices approach the ideal value 0.5. As a result, the weights will also asymptotically approach the value 0.5. This situation represents the fact that the dispersion in the distribution of employment of women is reduced over time. Thus the gender dimension is duly taken care of. It must be noted that 0<jt .
Now let IFt be the aggregate index of female financial empowerment across all the sectors of the economy for the time period t. This provides an overview of the status of female financial empowerment in the aggregative sense. Once [jt]s aresolved, IFt can be expressed as:
1t I1t + 2t I2t + 3t I3t
IFt =
1t + 2t + 3t
This multidimensional measure is a close variant of the Watt’s poverty measure (Tsui). Contrary to HDI (UNDP, 1990), here the weights are endogenously determined. The advantage is that, evaluation of the change in women empowerment between any two decades is free of any subjectivity. The changes in the magnitude and direction of inter-sectoral migrations of employed women are properly reflected in the estimation of changes in the financial empowerment of women over decades, when calculated with the help of the thus constructed index.
Cultural Empowerment: It is well recognized (Sen & Anand 2003, Mason 2005) that cultural empowerment is directly related to the level of education a person receives. Usually, higher level of education, higher is enlightenment and accordingly higher is level of empowerment one is entitled to enjoy in the society. Therefore, by looking at the level of education of women in the society concerned, at any given point of time, it is possible to measure cultural empowerment of women. Following the same procedure followed in case of financial empowerment, let Ijt be the percentage of girls in the total students enrolled in the jth sector at the tth period. [j=4,5,6 and t=1,2,…..,T]. Further let j=4 represent the primary level and similarly j=5 and 6 represent elementary and secondary levels of school education respectively.
Alike the previous case, here too, I4t, I5t, I6t are likely to be interdependent. The more the people realize that higher education is necessary for acquiring higher skill and this, in turn, is necessary for entitlement to better paid occupations, the more will be the incidence of enrolment in higher levels of education. Further, higher the incidence of enrolment of girls in the primary education, net of dropout, higher is the possibility of its counterpart in the elementary education. This in turn will imply possibility of higher enrolment of girls, in the secondary education. Thus enrolments at the above mentioned three tiers of education are vertically integrated.
As in the previous case of financial empowerment, characterization of this nature of inter-relationship could be expressed in the following fashion:
I4t = I5t5t I6t6t …(4)
I5t = I6t6t I4t4t …(5)
I6t = I4t4t I5t5t …(6)
For a given time period t, the above equations (4), (5) & (6) may be solved for a given set of I4t, I5t & I6t to obtain 4t, 5t6t.Here also, αjt represents importance or weight assigned to the jth tier of school education for the tth period. However, as usual, economic logic calls for 4t5t6t and 0<αjt. This fact again is provided by the very nature of interdependence of I4t, I5t & I6t. This means that the level of education where enrolment of girls is relatively less, should get higher weights and the vice versa. This is so because higher importance must be attributed to the higher level of education where the average level of knowledge or skill is higher. It may be noted here that higher skill provides women with more choice regarding their career, thus providing higher empowerment. Here also, along the path of development, dispersion in the enrolment of girls, across all the tires of school education, is likely to decrease over time. This implies, as in the previous case, the weights will asymptotically approach the value 0.5.
Once [jt]s are solved, an aggregative measure of cultural empowerment of women, denoted by ICt is expressed as:
4t I4t + 5t I5t + 6t I6t
ICt =
4t + 5t + 6t
Alike the measure of financial empowerment, this measure is also a close variant of the Watt’s poverty measure (Tsui, 2002). Here also, as usual, the weights are endogenously determined. The benefit is that, changes in cultural empowerment of women are objectively determined, accommodating proper reflections of inter-sectoral changes in the levels of education of women between the decades.
Aggregate Empowerment:By a close look, one can easily identify that education is the prime variable responsible for both the varieties of empowerment. Education imparts both skill and enlightenment. The former is responsible for financial empowerment whereas the latter leads to cultural empowerment. Thus interdependence between these two varieties of empowerment is natural.3