The Feminine Incidence Of Poverty

There is a widespread misunderstanding on the relationship between poverty alleviation and economic liberalisation. While one group feels that reforms will eradicate poverty, another feels that reforms will add to poverty. There is no theoretical, or empirical, literature in economics, which suggests that poverty will be eradicated with liberalisation alone. On the other hand, there is a huge body of economic literature which convincingly argues that poverty cannot be eradicated without economic liberalisation. Economic liberalisation, by increasing the national pie, enables everyone to get a larger piece. However, a larger size of the cake, does not necessarily imply that everyone gets more. Politically, it is easier to distribute the additional goodies to the have-nots, than to take resources away from the existing haves and then give them to the have-nots. In this sense, economic liberalisation is necessary, but not sufficient to alleviate poverty.

 

Those who want to stall economic liberalisation, or control it by giving a greater say to bureaucrats on where to liberalise, are missing the point altogether. Liberalisation, for sure, at least does not add to poverty. Moreover, it is the only politically feasible way to enable the government to eradicate poverty. The debate, therefore, should really be on whether the government is doing enough to distribute properly the additional pieces of the cake generated by the process of liberalisation. Stalling or modifying the liberalisation process, will be like throwing out the baby with the bathwater.

Lets look at some facts. Recently, under the sponsorship of the Department of Statistics, the Delhi Centre of the Indian Statistical Institute (ISI) completed a study to estimate the extent of poverty in the country. The distinguishing feature of the study is its attempt to measure poverty levels in every single NSS region (77 rural and 77 urban) for the NSS year 1987-88 and 1993-94.

At the aggregate level, there is a significant reduction in poverty, regardless of how it is measured, over these two years. Hence, there is no point trying to argue that poverty has gone up with liberalisation. One may say that one needs data for more years, but this line of argument cuts both ways.

What we concentrate on here is some of the other fallouts from this study, which can be of major significance. One such issue is the extent of gender bias in the incidence of poverty. Simply put, the question is: Are women more prone to poverty than men?

The gender bias can be at the workplace, or at home. If it is in the workplace, then governments need to enact, and enforce, laws that protect women at work. Here one can consider the efficacy of laws like job reservations of women. If the bias is at home, then such laws will have little or no relevance for the real problem.

At home, the gender bias may work in many ways. Compared to the boy child, the girl child may be given less nutrition, less education, less freedom to pursue her career, etc. This makes her less trained to enter the job market and, hence, prevents her from becoming economically independent of the man. Having job reservations for women will then either not work at all, or will reduce the short term efficiency of the economy at best. A better alternative will be to increase the awareness in the family, and ensure that education is received by all regardless of gender. This is admittedly more difficult than enacting laws regarding the marketplace. However, a responsible government should address the real issues rather than taking the easy way out.

To figure out whether there is a gender bias in the incidence of poverty, and how it operates, we carried out the following exercise. From the NSS quinquennial round expenditure surveys for 1987-88 and 1993-94, we identified each household by the sex of the head of the household. This was done separately for both the rural and urban sectors. Using the same poverty norm for both groups, we found that the households headed by women are more vulnerable to poverty than the households headed by men in the urban sector. In the rural sector, there was no significant difference in poverty incidence across the two groups.

Does this imply that there is no gender bias in the rural sector as far as poverty incidence is concerned? To resolve that, we have to first discuss the importance of the head of the household in this sort of exercise. Some maintain that the head in the NSS data is simply a reference point. Thus, the head is not necessarily someone with income earning responsibility. Consider a rural household where the adult male is the sole bread-winner but works as a migrant labour away from the village. According to the definition of an NSS household (people sharing the same kitchen), the migrant labour will not be a member of the rural household and, hence, will not be recorded as its head. A similar situation is less likely in the urban areas.

To grapple with this issue, we looked at women-headed households where the woman was currently married. The incidence of poverty of such households were similar to the men-headed households. On the other hand, when we looked at women-headed households, where the woman was widowed, divorced, or separated, the incidence of poverty was much higher than in the households headed by men. This then suggests that, where the female head of household is not simply a reference point (as in the household with the migrant male worker), the poverty incidence is higher than in male-headed households. Thus, the similar poverty incidence across male- and female-headed households in the rural sector was due to the significant presence of migrant labour households. Once we correct for this, women-headed households emerge more prone to poverty than men- headed ones.

Our next investigation was into the mechanism by which this gender bias operates. We calculated the poverty incidence by the sex of the head and the principal occupation of the household. We found that among households whose principal occupation involved skilled jobs, male-headed households performed better than female-headed households. We then checked for the educational levels (years of schooling) in the households, by occupational category and the sex of the head.

Women-headed households were significantly less trained than the men-headed households. Even within each household, the average male was better trained than the average female. Thus, one can hypothesise that women get screened out of skilled jobs, simply because they are not trained enough. The policy implication then: improve the educational levels of women to counter the gender bias in poverty incidence.

We have given here an example of what we think are the types of questions we must be asking to grapple with the problem of poverty. Valuable time is lost in taking ideological stances, which have little or no intellectual support.