Data Sources

Description and Limitations of Data

Statistics on Working Children

Statistics measuring the incidence of child labor in the individual profiles vary by age ranges and the definition used to measure child labor. There is no internationally endorsed definition on working children, or universally prescribed methodology for collecting data on child labor. The lack of concepts and methods for collecting child labor data has made it difficult to obtain comparable and reliable statistics across countries on working children. Therefore, estimates on the number of working children presented in this report come from government household surveys and international agencies that use varied age ranges and different definitions to construct child labor statistics. In general, estimates on the number of working children are likely to be underestimates because the nature of household surveys do not lend themselves to collecting data on children who are working in the informal or illegal sectors of the economy, particularly children in the worst forms of child labor.

Resources used in this report for child labor statistics that use household surveys such as national census or labor force surveys come from ILO’s Yearbook of Labor Statistics, ILO’s Economically Active Population Estimates and Projections: 1950-2010 database, and World Bank’s World Development Indicators. Estimates on the number of working children from these sources are generally based on the definition of the “economically active population” which restricts the labor force activity of children to “paid” or “unpaid” employment, military personnel, and the unemployed. The definition does not include children in informal work settings, non-economic activities, or “hidden” forms of work such as domestic service, prostitution, or armed conflict.

Other sources on child labor statistics come from ILO-IPEC’s SIMPOC and UNICEF’s Multiple Indicator Cluster Surveys (MICS) that are specifically designed to measure the extent of child labor in a country. Although these surveys also employ the definition of “economically active,” the universe for working children is extended to include work activities outside the definition, which these surveys use to capture children working in the informal sector and domestic work. However, the methods for collecting data on the worst forms of child labor adequately capture the distinction between the worst forms and other forms of child labor.

Another main source used in this report is from country statistics on child labor provided by the Inter-Agency Research Cooperation Project on “Understanding Children’s Work” (UCW) from the ILO, UNICEF, and the World Bank. The UCW project has compiled statistics from approximately 50 countries that use both traditional household surveys and child labor surveys. Estimates on the number of working children from this source also vary by age ranges and definitions.

In determining which source to use in presenting estimates on child labor for the individual profiles, priority is given to statistics collected from child labor surveys such as the ILO-IPEC SIMPOC survey or the UNICEF MICS survey. In instances where data is not available specifically from a child labor survey, estimates based on household survey data from the UCW project are presented to reflect the number of children working, children who combine school and work, and children who work in household chores for a specified amount of time. Finally, countries that do not have data on working children from a child labor survey or the UCW project draw on estimates from ILO’s Yearbook of Labor Statistics, ILO’s Economically Active Population Estimates and Projections: 1950-2010 database, or the World Bank’s World Development Indicators. In some cases, recent statistics from child labor surveys do not exist, and estimates from other more recent sources are presented along with estimates from the child labor survey.

Statistics on Primary Education

In addition to data on the number of working children, statistics on primary school attendance or enrollment are used in this report to provide a complementary indicator of the number of children who work. The presentation of primary attendance statistics allows the reader to infer the proportion of children in the school-age population who are not in school and may be engaged in child labor or at risk of entering hazardous work activities. While primary attendance statistics are more accurate than primary enrollment statistics in illustrating the extent of child labor in a country, attendance statistics are not often collected and readily available. For countries in which primary attendance statistics do not exist, primary enrollment statistics are used instead.

There are, however, several limitations to using primary enrollment statistics that, when presented along with child labor data, can be misleading. Since child labor data and education statistics are usually collected separately and for different purposes, data on the percentage of children enrolled in school may not necessarily be consistent with statistics on working children. Primary enrollment statistics measure the number of children who are enrolled during a given school year, but do not reflect the number of children actually attending school. Thus, a child can be enrolled in school, but never attend. As a result, primary enrollment statistics often overstate the true number of children who attend classes on a regular basis, and underestimate the number of children who may be working. In other cases, children may be enrolled in or attending school, and are also engaged in a worst form of child labor outside of school hours.[28] Nevertheless, to the extent that child labor and education are linked, it is still important to present some sort of education statistics because primary enrollment and attendance figures provide a measure of access to quality and relevant education for children, particularly those living in rural areas.

Data Sources on Child Labor and Primary Education

Sources of Child Labor Data

Child labor statistics in this report were obtained from the following four main sources:

ILO-IPEC Sponsored Child Labor Surveys

Since 1992, ILO-IPEC and ILO-STAT have worked in close collaboration with national statistical agencies and other institutions in several countries to design and implement specialized surveys on child labor.[29] Eleven child labor surveys were conducted under the direction ILO-STAT, and administered either as a stand-alone survey or a child labor module as a part of a country’s existing labor force surveys. In 1998, ILO-IPEC launched SIMPOC in which 34 countries either have completed or are in the process of collecting child labor data.[30] The population of working children generally includes children between the ages of 5 and 17 who are employed, unemployed or domestic workers in their own household. Children who are working are either salaried, self-employed, unpaid family workers, or apprentices. Unlike traditional labor force surveys, the SIMPOC child labor surveys collect data on some work activities in the informal sector.[31]

UNICEF’s Multiple Indicator Cluster Surveys (MICS)

In 1998, UNICEF began a process to assist countries assess progress in meeting their goals for the “World Summit for Children” using the MICS questionnaire.[32] The MICS questionnaire includes 75 indicators for monitoring children’s rights such as child labor, child survival and health, child nutrition, maternal health, water and sanitation, and education. Child labor measures consist of children between the ages of 5 and 14 who are paid, unpaid, or work at least 4 hours a day in domestic work.[33] About 49 developing countries included the child labor indicator in their MICS questionnaire;[34] to date, 28 countries have submitted their national reports to UNICEF.[35]

The Inter-Agency Research Cooperation Project on “Understanding Children’s Work”

The “Inter-Agency Research Cooperation Project: Developing New Strategies for UCW” is a group collaboration between ILO, UNICEF and the World Bank to minimize the duplication of efforts on child labor statistics among the three agencies, and identify information gaps in the data on child labor. ILO contributes data from its SIMPOC surveys, UNICEF from the MICS questionnaire, and the World Bank from its Living Standards Measurement Survey. The UCW project has country statistics on child labor, schooling and health indicators from government census or household surveys sponsored by at least one of the three agencies.[36]

Estimates and Projections of the Economically Active Population, 1950-2010

Child labor force participation rates are taken from the World Bank’s World Development Indicators 2003 or the ILO’s on-line database for labor statistics, which are based on data from the ILO’s database Estimates and Projections of the Economically Active Population (EAP): 1950-2010. Statistics from the ILO’s EAP database are different from statistics from ILO’s SIMPOC child labor surveys. Labor force participation rates from the EAP database are based on the definition of the “economically active population” for children between the ages of 10 and 14.[37] Although the EAP is less accurate in estimating the number of children working below the age of 15, it does provide a useful indicator because it is the only available source for comprehensive and comparable data on working children 10 to 14 years old.[38]

Sources of Primary School Education Data

Primary school education data for gross and net primary school enrollment were obtained from either the World Development Indicators 2003 or the Education For All 2000 Assessment. Gross and net primary school attendance rates were mostly obtained from USAID’s Global Education Database.

World Development Indicators 2003 (WDI 2003)

The WDI 2003 is an annual compilation by the World Bank on development data gathered from several international and government agencies and private and nongovernmental organizations around the world. The WDI 2003 includes 800 indicators on topics in six areas: world view, people, environment, economy, states and markets, and global links. There are 85 tables covering the six categories with basic indicators on 224 countries.[39]

Gross and net primary enrollment statistics in the country profiles primarily use data compiled in the WDI 2003. While the Department of Labor’s 2002 Findings on the Worst Forms of Child Labor used similar 1998 data compiled in the WDI 2002, statistics presented in this year’s report using data for the same year from the WDI 2003 may differ slightly because of statistical adjustments made in the school-age population or corrections to enrollment data. In several instances, there was no change in results from the WDI 2002 and WDI 2003 because enrollment statistics were not affected by the adjustments or corrections to the data were not needed.

Education for All 2000 Assessment

The Education for All (EFA) Year 2000 Assessment Statistical Document is a collection of quantitative data based on an in-depth evaluation of basic education at global, regional, and national levels gathered by UNESCO Institute for Statistics (UIS). Data is available for 185 countries. Indicators examined in this assessment include the demand for education, early childhood education and care, access and trends to primary education, public expenditure on primary education, teacher qualifications, and literacy rates.[40]

Global Education Database 2000 (GED)

The GED provides education data compiled by UNESCO and from USAID Demographic and Health Survey (DHS) sources. Data include education measures on the performance of a specific country or groups of countries over time in areas such as school enrollments and attendance, public expenditure, and gender parity. For most of the 145 UNESCO indicators, data are for 1980, 1985, and for single years from 1990 through 1997 or 1998. DHS statistics are presented for the specific country and year in which surveys were conducted and are reported through 1999. With over 200 countries represented, the database is a useful tool for cross-country comparisons of education indicators.[41]