|Keywords (Extracted from title, table of contents and abstract of thesis)
income and consumption inequalities, generalized entropy indices, gini coefficient, generalized gini indices, atkinson's indices, ebertā€™s indices, earned incomes, non-earned incomes, inequality measure
This thesis presents an overview of alternative approaches to measuring inequality and analyzes various dimensions of income and consumption inequality in Pakistan using alternative approaches. It is observed that generalized entropy indices, Gini coefficient, generalized Gini indices, Atkinson's indices and Ebertā€™s indices pass the tests of desirable properties of an ideal inequality measure. Among these the generalized entropy indices, Gini coefficient and Ebertā€™s indices are also decomposable. Some of the popular measures like the two Theil's measures, Gini coefficient and Atkinson's indices can be derived from these general measures either as special cases or as transformations of the special cases. The thesis employs all these measures for the empirical analysis. Inequality in income and consumption is analyzed using a) households income and household consumption; b) per adult-equivalent income and consumption; and c) earnings of individual earners. The entire analysis is based on micro level data taken from Household Integrated Economic Survey for the years 1992-93, 1996-97, 1998-99 and 2001-02.
The study shows that in general inequality in household income/consumption has been greater than the inequality in per adult-equivalent incomes/consumptions, indicating that ignoring household size tends to overestimates the extent of inequality. The level of consumption inequality is generally observed to be less than the level of income inequality. The analysis further reveals that income and consumption inequalities have been more severe in urban areas than in the rural areas. This difference can be explained by the prevalence of institutional wage rewards in rural areas whereby almost all members of labor force are employed but the wage rate is set around the subsistence level and high rents and transportations costs in urban areas. It is observed that the inter-regional inequality in income or consumption is mainly across the rural -urban divide rather than across the provincial divide. The analysis further shows that the rural -urban inequality is higher in the larger provinces as compared to the smaller ones.
Decomposition analysis of income shows that non-earned incomes are more unequally distributed than the earned incomes. The sources of income that have contributed significantly to inequality are self-employment other than agriculture, profits, interests, gifts/assistance and property in urban areas and self-employment in agriculture as well as other than agriculture and property income in the rural areas. The analysis further shows that inequality in non-food consumption expenditures is more severe as compared to the inequality in food consumption expenditures.
The thesis finds the presence of high degree of earnings inequality in Pakistan, especially in urban Pakistan. The analysis further shows that earnings inequality is relatively higher among youngest and eldest groups of earners, female earners, earners with low level of education, earners working as employers and members of producer's cooperatives, and the earners employed in services industries other than social & personal services. On the other hand, the degree of inequality is lower among the middle-age groups, the highly educated workers, paid workers and the workers employed in the primary products industries namely agriculture, forestry, hunting & fishing.
The decomposition analysis indicates that age of the earners in an important factor contributing to earning inequality in urban Pakistan but not in rural Pakistan. In urban Pakistan the earnings inequality between employment statuses, especially between employers and paid employees, is also quite high. Contrary to the popular belief, the inequality between earners with different levels of education is not much high, though the inequality within the groups of earners having higher levels of education is less than the inequality within the groups of earners with low levels of education.