Printer Friendly

II) Data and methodological approach.

The data used in this paper are those collected as a part of the Demographic and Health Surveys (DHS). These are large, nationally representative household surveys, and the data from 57 surveys (from 41 countries) are analyzed here. (2) Basic information on the number of households in each sample, as well as the number of individuals in the sample of 6 to 14 years olds, and 15 to 19 year olds, are in Table 1. The DHS were not designed to collect information on education. Rather, they were a systematic data collection effort whose main purpose was to obtain nationally representative and cross-nationally comparable household-level data related to family planning, and maternal and child health. The more recent surveys did record data on school enrollment (for household members aged 6 to 25) and educational attainment (for household members aged 6 and above) as reported by a chosen respondent.

Data on education outcomes

The education variables analyzed here are based on the answers to three questions about those aged 6 and above: whether they had ever been to school; if they had ever been to school, what was the highest level of schooling attended; and what was the highest grade attained at that level. Those aged 6 to 25 were asked, in addition, whether they were still in school" (if they report ever attending). In the rest of this paper, children who report being in school" are referred to as being enrolled.

The countries have been grouped into eight regions for the analytical purposes of this paper. These are, ranked roughly from lowest to highest enrollment of girls aged 6 to 14 from the poorest households: Western and Central Africa, North Africa, South Asia, Eastern and Southern Africa, Central America and the Caribbean, East Asia and the Pacific, South America, and Middle East and Central Asia.

Measuring wealth using DHS data

The DHS do not ask about household income or consumption expenditures, the variables usually used to rank households according to their standard of living. The surveys carried out since 1990 do however include two sets of questions related to the socio-economic status of the household. (3) First, households are asked to report about ownership of various assets, such as whether any member owns a radio, television, refrigerator, bicycle, motorcycle, or car. Second, questions are asked about housing characteristics, namely whether electricity is used, the source of drinking water, the type of toilet facilities, how many rooms there are for sleeping, and the type of materials used in the construction of the dwelling. There is substantial overlap in the questions asked in different countries, but the precise list varies. The number of variables derived from these questions is usually 15 or 16 but varies from 9 to 21 (shown in the last column of Table 1). (4)

[Table 1 about here]

In order to use these variables to rank households by their economic status, they need to be aggregated into an index, and a major problem in constructing such an index is choosing appropriate weights. (5) This is done here using the statistical technique of principal components. Principal components is a technique for summarizing the information contained in a large number of variables to a smaller number by creating a set of mutually uncorrelated components of the data. Intuitively, the first principal component is that linear index of the underlying variables that captures the most common variation among them.

The details of the methodology are described and defended in Filmer and Pritchett (1998) which shows that the asset index performs as well as a more traditional measure, such as household-size-adjusted consumption expenditures, in predicting educational enrollment and attainment. (6) The methodology was applied in Filmer and Pritchett (1999a) to analyze wealth gaps in educational attainment in 35 countries, and in Filmer and Pritchett (1999b) which investigates the determinants of education gaps in India, and how these vary across states. This paper extends these previous analyses by highlighting how gender interacts with wealth, adult education and the presence ofschools in the community, and how these relationships differ across countries and regions.

The fourth column of Table 1 shows how well the first principal component of the asset variables (which is the asset index) "fits" the underlying variables, reporting the proportion of the variation captured. The proportion is remarkably stable, and reasonably high, at between 20 and 30 percent of the variance (ranging from Malawi, Tanzania and Uganda at 19 percent to Bolivia at 31 percent). (7)

The asset index is calculated separately for each country. Within each country individuals are sorted by the asset index, and cutoffs for the bottom 40 percent, the middle 40 percent, and the top 20 percent of the population are derived. Households are then assigned to each of these groups on the basis of their value of the asset index. (8) From here on these groups are referred to as "the poor," "the middle," and "the rich". Reference to a "poor" child should be read as "a child from a household in the group in which 40 percent of the population with the lowest asset indexes live."

A note of caution is warranted here: the principal components procedure normalizes the mean of the index to zero for each country. Therefore, when comparing the "poor" in Kenya to the "poor" in Turkey or India it is important to keep in mind that the measure is relative, and 40 percent of the individuals are defined as living in "poor households" in every country. This paper does not attempt to generate an absolute poverty measure based on the asset index approach. (9) As a rough benchmark, Table 2 reports the percentage of the population living below the national poverty line, the dollar-a-day and the two-dollar-a-day poverty lines for the countries analyzed here as reported in the World Banks World Development Indicators database (World Bank, 1999). The percent who live below a dollar a day clearly varies tremendously across countries, from below two percent in Morocco to almost 90 percent in Haiti. In an (unweighted) average across these countries the percentage living below this internationally comparable poverty line is about 40 percent--the percentage defined as the 'poorest group" in the analysis in this paper. National poverty lines produce a much more stable proportion of each country defined as poor: again, the cross-country (unweighted) average is again about 40 percent.

[Table 2 about here]

What to take from this? Although using an asset index approach does not provide an internationally comparable cutoff (in the sense that a dollar-a-day day does) it does identify a group of individuals in each country whose size is comparable to other breakdowns that are frequently made. In particular, using the 40 percent cutoff in this paper corresponds approximately to the percentage of people living below the national poverty line in many countries (Cameroon, Bangladesh, India, Nepal, Kenya, Philippines) or the percentage living under a dollar-a-day in Zimbabwe two-dollars-a-day in Brazil. (10)
COPYRIGHT 1999 The World Bank
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1999 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:The Structure of Social Disparities in Education: Gender and Wealth
Author:Filmer, Deon
Publication:The Structure of Social Disparities in Education-Gender and Wealth
Date:Nov 1, 1999
Previous Article:I) Introduction.
Next Article:III) The magnitude of gender and wealth differences in enrollment.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters