[Excerpt] At the present time, there is great interest among development economists in the problem of economic inequality in less developed countries (LDCs). Studies of the determinants of inequality follow either of two general approaches. The more traditional approach is associated with names like Kuznets (1963), Chenery and associates (1960, 1968, 1975), Adelman and Morris (1973), Ahluwalia (1976) and Chiswick (1971). These studies share a common methodology, consisting basically of looking at a cross-section of countries, and (1) measuring the degree of inequality in each, (2) measuring other characteristics of each country (e.g., level of GNP, its rate of growth, importance of agriculture in total product, etc.), and (3) relating the level of inequality to that economy's characteristics using correlation or regression analysis.
In the last few years another type of approach has been followed, which looks instead at inequality within a country, and measures the contribution of the various components to total inequality. In this type of approach, using a variety of methodologies, inequality has been decomposed by economic sector (e.g., urban vs. rural), income source (e.g., income from labor vs. capital vs. land vs. transfers), or family characteristics (including attributes of the workers, their jobs, and regional and other locational considerations). This mode of inquiry is potentially of great value for understanding the structure of inequality and identifying which are the most important explanatory factors.
This study explores the decomposition type of inequality analysis. I summarize the alternative decomposition methodologies which have been set forth in the literature and review the principal findings of empirical studies.