TITLE: Decomposition of Consumption Inequality in Ecuador

AUTHOR: Xavier Arcos C., MPS-International Development, Cornell University

ADVISORS:
Prof. Gary S. Fields, Department of Economics, Cornell University
Prof. Lourdes Beneria, Department of City & Regional Planning, Cornell University

The paper attempts to determine the characteristics of consumption inequality in Ecuador as of 1994. The research was possible thanks to recently developed methodologies and data available from the 1994 Ecuador Living Standards Measurement Survey. The paper is divided in two sections which address from different perspectives the topic.

The first section contains a decomposition of consumption inequality between groups and within groups. Two exercises where carried out: the first dividing the population by rural and urban sector; the second dividing the population by sex of the head of the household. The purpose is to determine whether inequality is given by: 1) one group being wealthy and the other being poor; or 2) having both rich and poor people within each group. For example for the sectoral exercise it is possible to determine whether inequality is given by a concentration of rich people in the urban sector and poor people in the rural sector, or whether inequality is significantly present in both sectors individually. The relevance of this kind of analysis is that it may affect policy decisions about directing more resources to the rural sector versus urban sector.

The results indicate that of total inequality only 19% accounts for inequality between urban and rural sector. In other words there are differences between both sectors with the urban sector being better off, however most of inequality occurs within each of the sectors rather than across them, therefore, it is not true that poor people are concentrated in the rural sector and rich people are concentrated in the urban sector. Hence development policies in Ecuador should be directed toward targeting this kind of situation instead of toward reducing inequalities between sectors.

On the other hand, of total consumption inequality, 99.6% accounts for inequalities within each group of male and female head of the household. This indicates that there are no differences between both groups, most of inequality is present in each group individually. Therefore, there is no need for development policies directed toward targeting the population by sex of the head of the household. Note that this analysis refers only to the head of the household, so by no means it can be generalized to all the male and female population. In fact, one of the recommendations for future research is to repeat the exercises carried out in this paper using information of future surveys which contain consumption or income data at personal level by sex.

The second section analyses a decomposition of consumption inequality by consumption determinants. The purpose of this section is to determine what percentage of inequality is explained by variations in levels of education, years of experience, sex of the head of the household, size of the household, and sector. Policy makers could direct their attention and efforts to an specific field in order to effectively deal with inequality by identifying the main inequality determinants. The results indicate that of total consumption inequality, 31% is accounted by variations in education, 17% is accounted by variations in the size of the household and 9% is accounted by variations in the sector. Sex of the head of the household is not relevant in explaining inequality. Years of experience on the other hand actually contributes to equality, in other words the negative sign of the coefficient for experience indicates that it contributes to the reduction of inequality. The five variables considered together account for 51.6% of variations in consumption.

The above leads to conclude that when looking at inequality policy makers should pay special attention to the factor education and size of the household. Variations in education seem to explain most of the variations in the levels of consumption, followed by the factor size of the household. Thus the implementation of aggressive policies directed toward making education accessible to the majority of population plus birth control programs should in the long term contribute to reduce inequality more efficiently than any sectoral policy.

The result for the variable size of the household is intuitive, the more children a family has, the less the consumption per capita. Finally, sector does not seem to be of particular importance since it explains only 9% of inequality, thus sectoral policies directed specifically to allocate more resources in the rural sector than in urban sector should be discouraged. Note that this is consistent with the result found on the first section of the paper where a sectoral decomposition of consumption inequality was performed.

To order a copy of the paper in the U.S. and Europe send $6.00 for postage & printing costs. In Ecuador send $3.00. Include your complete mailing address. Write to:

Xavier Arcos C.
Manuel Barreto 239 y Coruņa
Quito, Ecuador

Phone: (5932)528-943, (5932)554-227

Email: xra1@cornell.edu

Go back to Ecuador Economy Homepage