The following tables present data relevant to the economics of child labor. Tabulations of household level survey data collected in nine countries under International Labor Organization (ILO) sponsorship are presented, as well as two country tables produced using data from the World Bank’s Living Standards Measurement Survey (LSMS) project. Nepal is represented in both the ILO and World Bank tables. 

Because each country participating in either the ILO or World Bank programs have conducted their data collection efforts independently, percentages are not directly comparable across countries. Data definitions, concepts and culturally driven perceptions of these concepts can cause differences in measurements of child labor across countries that would not exist if the data were generated using the same definitions and concepts, or if concepts were viewed through a common cultural filter. Indeed, this issue of comparability can even sometimes hamper comparisons within different areas of the same country. In Indonesia, for example, the concept of work varies among different regions:

The percentage of respondents to the word “working” is also dependent on cultural values. For instance, if the question of working is asked in Bali the answer is almost certainly “yes,” indicating that everybody is working. This is due to the fact that Balinese see any activity as a working activity. But in other parts of Indonesia, for example in Yogya, the word “working” applies only to those who work as civil servants.¹

As a result, figures on the incidences of child labor should be considered not to be directly comparable across countries. Instead, identifying common patterns or trends is a more appropriate objective. 

Defining “economically active” work was not an easy task for the statistical authorities collecting the data. The standard method for defining economically active work is to consider tasks done for an enterprise outside the family as economically active, whether or not children are paid. Tasks performed within the home for one’s own family are also considered “economically active” if the products of such work contribute to household income. Household chores that simply benefit the family—such as caring for one’s siblings or producing food for the family’s own consumption—however, are not typically considered “economically active” work. Countries which differ from this standard definition are noted within footnotes at the end of the individual tables.

Tables for each country are divided into two to four panels. Panel A contains economic activity rates for a variety of subgroups of children. Panel B compares economically active and non-active children along specific dimensions. For example, the third entry in Panel B in Table B-1 gives for Bangladesh the relative frequency of various levels of educational attainment for active, non-active, and all children; and shows that active children are disproportionately more likely to have no education at all (88 percent of active children have no education, versus ten percent of non-active children and 25 percent overall). Panel C gives an indication of the type of work in which active children are engaged. Panel D (available only for the Philippines in Table B-7) gives reasons why children work.

It is important to note the difference between “usual” and “current” definitions of economic activity. Those surveys which measure economic activity according to the “usual” definition of economic activity consider children who work, or look for work, for at least one hour in the twelve months prior to survey as “economically active.” Surveys that measure children according to the “current” definition of economic activity consider children who work or look for work for at least one hour in the week prior to survey as “economically active.” Most often, figures presented here are based on the “usual” definition. Variations from this standard are noted where relevant.

Most country studies are nationally representative, even when figures are based on a relatively small sample. Ghana and Indonesia provide exceptions to this rule. In these two countries, samples on which the numbers are based are not representative of the country as a whole. Instead, these countries sampled households within only a select couple of regions.

Finally, the data provided here are only a “snapshot” of the child labor situation in the countries represented. They are suggestive of factors or variables that may be related to child labor, either as causes or outcomes, but they are not conclusive find- ings. Ideally, these data should serve as a guide to further analyses that seek to relate child labor to a variety of variables—not just one or two at a time—so that the interaction among variables and the relative importance of each can be assessed. Critical to these analyses is understanding the decision making process that leads households to send their children to work. 

The next step for data analysis is to look at how households with varying characteristics differ in their child labor decisions. Analysis based on statistical techniques such as multiple regression is needed, and for this, data records for each house-hold surveyed are the necessary input. At the moment, such data are available to researchers from the LSMS program.² ILO sponsored data are currently available only in tabulated form. However, the ILO has plans to sponsor new household level surveys in over forty countries, and it is expected that household level survey data from many of these surveys will be available to researchers for rigorous analysis. Thus the ability to go beyond tabulations such as those presented here, and accordingly to draw firmer conclusions about the causes of child labor in a variety of countries, should be enhanced in the future.

¹ Working Children in Indonesia (Bandung: Central Bureau of Statistics, 1993) 39.

² More information on the Living Standards Measurement Study may be found at: www.worldbank.org/lsms. For many
countries, household level data and necessary documentation may be downloaded directly. Other countries’ data and
documentation require a written request (and sometimes a fee payment) to the particular country and to the LSMS office.
In most cases, the household level is available once these permissions are granted.”