Click here for search results


This site presents "profiles" of education outcomes for many developing countries, as derived from an analysis of various household survey data. The idea behind these pages is to let you navigate through the different profiles and discover the patterns in educational outcomes and how they vary across countries and groups within countries. If this is your first visit to this site please read the examples below to help make your browsing more productive. To select a data set, or several data sets, go to the main selection page.

ProfilesPopulation | Data Sources | References

methodsFor example, you will be able to find the proportion of 15 to 19 year olds who have completed grades 1, 2, 3, and so-on, or the proportion of children of each age between 6 and 14 who are currently reported as being in school. Moreover, you will be able to compare these profiles across subgroups of the population (e.g., economic status, males and females).


Five main types of profiles can be generated: 

A ttainment profiles for the current cohort

methodsThese figures show the proportion of the cohort of individuals currently between the ages of 15 and 19­inclusivewho have completed each of the grades.

The figures enable one to see, not only the aggregate proportion of the particular cohort of 15 to 19 year olds who have completed a grade, but allows the synthetic decomposition of that proportion between those who "ever enrolled", and those who left the school system at each grade (synthetic because the figures do not track individuals over time, but are averages for the cohort).

In Benin, for example, only 26 percent of the those aged 15 to 19 from the poorest 40 percent of households (as ranked by an index of household asset ownership) had completed grade 1 or higher, meaning 74 percent had never attended school or not completed even one year of schooling. Only 7.9 percent had completed grade 5 or higher, and only 0.7 percent had completed grade 9 or higher.

Among the richest 20 percent of households, 80 percent of 15 to 19 year-olds had completed grade 1 or higher but drop-out was such that only 54 percent had completed grade 5. Even among the rich only 17 percent had completed grade 9.

Back to profile gif | 

Attainment profile for different cohorts

These figures show the attainment profiles, not only for the cohort of those aged 15 to 19, but also for the cohort of those currently 20 to 29 years old, and those 30 to 39 years old. This view allows one to see how attainment patterns for a given country have changed over time.

For example in Bangladesh, in 1996-97, 77 percent of males currently aged 15 to 19 had completed grade 1, and 69 percent of females had; only 26 percent of females currently aged 15 to 19 had completed grade 8, but among those currently 20 to 29, only 20 percent had, and among those 30 to 39, only 12 percent had; for males, the percentages were 32, 34, and 28 percent.

Back to profile gif |

Cross sectional enrolment profiles, Nepal, ages 6-14Cross-sectional enrollment profiles

These figures show the proportion of children of each age (from 6 to 14) who are currently enrolled in school (the numbers are derived not from official records but from answers to a question such as "is this child still in school?").The figures enable one to see the ages at which children tend to enroll in school, and subsequently drop out.

For example, in Nepal, in 1996, only 55 percent of girls aged 7 are currently reported as "in school" whereas 75 percent of 7year-old boys are; for boys, 84 percent are currently enrolled at age 11 and for girls 48 percent are currently enrolled at age 13.



Back to profile gif |

 Cohort grade survival profiles

 methodsThese figures show the proportion of children of the cohort of 10 to 19-year olds who have completed each grade (including grade 1) each age, using techniques adapted from "survival analysis." The technique estimates the proportion of the cohort that has completed a gradeadjusting for the fact that some students are still in school and we cannot observe their ultimate grade completed (the observation is said to be "right censored"). The figures enable one to see the estimated grade-to-grade progression of a current cohort.

For example in Angola, 84 percent of the current cohort of children (ages 10 to 19) has survived to grade 1, and and estimated 53 percent have survived to complete grade 9, adjusting for those that are still in school and may yet complete grade 9.




Back to profile gif |

Proportion of population in each level of schooling, ages 6-24, Nepal (corrected)Enrollment pyramids

These figures show the proportion of 6 to 24 year olds who are currently enrolled in school, and specifies the level of schooling they are enrolled in.

For example, in Nepal, in 2001, while 79 percent of 7 year-old boys were enrolled in primary school, only 66 percent of 7 year-old girls were; and 42 percent of 14 year-old boys were enrolled in primary school, while 33 percent were enrolled in secondary school.


Back to profile gif |

 Population subgroups

These profiles can be presented for all or different subgroups of the population:

  • By economic status. These are for individuals from households in the poorest 40 percent, the middle 40 percent, and the richest 20 percent as defined by an index based on the ownership of various assets by members of the household, the "asset index" [1]. The economic status groups are redefined for each country so that the groups are relative within countries. That is, a household defined as 'poor' in one country might well be considered 'rich' in another country. Since household economic conditions at the time of schooling of older cohorts (those aged 20 to 29, and 30 to 39) are less likely to be well captured by the current measure of economic status, the profiles for these cohorts are not disaggregated by economic status.
  • By economic status quintiles. These are simply a more disaggregated version of the economic status groups with individuals grouped into the poorest 20 percent, the next 20 percent, and so on.
  • By household per capita expenditure groups or quintiles. These use the more common consumption expenditure based measure of welfare to rank individuals from poorest to richest. This measure in not available for the DHS or MICS2 surveys
  • And by gender,   urban/rural residential location,  combination of the gender and location groups, and combination of the gender and economic status/per capita expenditure groups.

Data sources

The profiles presented here are derived from the analysis of data from the following sources:

All of these data sets are large household surveys that aim to be nationally representative (although in some cases, parts of countries are excluded when surveys could not be implemented in certain regions).

The education variables analyzed are typically based on four questions:

  1. Has a household member ever been to school?
  2. If so, what is the highest level of schooling attended?
  3. What is the highest grade completed at that level?
  4. Is this household member still in school?

The figures presented on this site are the result of the analysis of these household data sets. As such, they represent outcomes at the time of the survey and are not always representative of the current state of education in a given country.


[1] Estimating Wealth Effects without Expenditure Dataor Tears: With an Application to Educational Enrollments in States of India(Filmer and Pritchett, 2001).

Updated: July 8, 2009

Permanent URL for this page: