It is evident that answering all the above questions would require the collection of comprehensive information on working children. Consequently, the ILO Bureau of Statistics designed four survey approaches and tested them in a number of countries together with a supplementary inquiry. Three of the survey approaches were implemented, respectively, at the level of households, employers/establishments/ enterprises, and street children. The fourth method tested was a "rime use" approach. The supplementary inquiry was applied at the community level (cities, towns, villages). The main purpose was to determine which survey methodology would yield the best results.
The surveys measured as many variables as possible, particularly in relation to the various non-schooling activities of children in the 5-14 age group, their characteristics and those of their parents or guardians, and so on. The principal variables considered for the investigation related to the following subjects, as expressed in broad terms:
Definitions of child labour
The concepts, definitions, classifications and the like, used for the purposes of the experimental surveys in all the countries, were generally in line with internationally recommended standards concerning such elements as the economically active population, the labour force, classifications of industry, occupation, status in employment, age grouping, households, enterprises and establishments and so on (see Chapter 2), with some variations to reflect the unique circumstance of child work and the peculiarities of the individual countries.
Depending on the availability of basic information or demarcations regarding the general characteristics of the areas covered and the availability of appropriate sampling frames, the different elements considered for the stratifications included development levels of the selected rural and urban areas - for example, poorly developed/well developed, slum/non-slum blocks, income classes (low, middle, high), overall rates of literacy/illiteracy of the general population, school attendance levels and so on. This was because it is known that factors such as these and the incidence of child labour are either positively correlated or van- inversely, depending on the factor being considered.
For the purposes of the experiments, a "child" was defined as a person between five and 14 years of age. In the absence of a universally endorsed definition of "child labour", all activities of children were enumerated and quantified so that the data could be tabulated according to the different characteristics or categories of the variables included in the questionnaires. Depending on the level and nature of the quantified activities or variables, those which were judged or expected to have negative effects or consequences on the health, education and normal development of the working child were considered as falling within the boundaries of "child labour".
The main focus of the surveys in all four countries was on the economic activity of the children, whether paid in cash or in kind, or in unpaid family work, thus respecting the international definition of "economic activity". In this respect, some types of work or production for own household consumption, such as carrying water, fetching firewood, pounding and husking food products, were also considered as falling within the boundaries of economic activity. While the dividing line between economic and non-economic activities for cases such as the above is rather thin and not always obvious, these and many others (for example, preservation of fruit by drying or bottling, weaving cloth, and dressmaking and tailoring), were considered to fall within the margin of economic activity or the "production boundary" as defined by the System of National Accounts (SNA, 1993).
While schooling activities were measured in the majority of cases, in some instances non-schooling activities of a non-economic nature (especially household chores or housekeeping services provided in the child's own parents' or guardians' homes) were also estimated separately. In all the surveys, both the "current" and the "usual" economic activity approaches were applied, the first in reference to activities during a short period such as the week (or seven days) prior to the date of the interview, and the second in reference to a long period such as the 12 months (or 365 days) preceding the inquiry date. The latter reference period takes seasonality into account, which is an important factor since a considerable proportion of children's activities is seasonal, including activities undertaken when schools are closed.
In all the selected areas the household-based surveys were carried out strictly on a relatively rigorous sample basis using a multi-stage (two- or three-stage) stratified sampling design. Using the household listing as a sampling frame as well as the basic information that was collected during the listing, all the listed households in each unit of the segment were then grouped into three strata as follows:
As a final step in the sample selection procedure, a specified number of households in each of the above three strata were selected by means of systematic sampling which formed the final stage sampling units. Through these sampling procedures or slight variations, between 4,000 and 5,000 households were selected to represent the sample size for the surveys in each of the four countries.
Where suitable statistical software packages were not available in the statistical offices of the countries concerned, a self-weighting systematic sampling design with probability proportional to size (PPS) was adopted. This approach helped by providing a uniform weight for estimating totals. It also facilitated the computation of percentages, means and ratios of the population parameters directly from the sample data.
The questionnaire that was applied at the household level consisted of two parts. The first part was addressed to the head of the household (or a proxy) to obtain information on the demographic and socio-economic composition of the household, including such aspects as housing facilities, household migration status and living standards, and the education level and economic activity status of the household members. The second part was used to collect the required information from the individual children themselves.
As a supplement, a simple questionnaire was used to interview elected and appointed leaders, administrators and the like, in the communities or towns and villages of the selected areas so as to identify the major local socio-economic characteristics, assess development levels and determine the differential in the incidence of child labour. This investigation was also used to list the households which served as a sampling frame.
The details of the variables considered for the household-based survey approach and the community/town/village level, as well as those for establishments or employers, are listed in Appendix 3.1.
During the household listing stage, basic information on a limited number of variables relating to each household and its members was also obtained to facilitate the stratification of the households in each segment and the sample selection of households.
The employer's (establishment or enterprise) questionnaire was addressed to the owner of the business or a designated respondent, seeking information on the particulars of the ownership, the goods produced and services rendered, the number of children and adults engaged, their working conditions, the reasons for using child workers, facilities and health care at the workplace and so on.
In this approach, probability sampling became prohibitive due to the absence of basic information which could serve as a master frame, such as an exhaustive list or directory of employers in respect of the areas selected for the surveys. In view of this problem, only those employers identified by the children themselves or their parents during the interview at the household level, or those enterprises known or suspected to be using child workers, were located and interviewed on a random basis. In this way, up to a total of 200 entities were identified and enumerated in urban and rural areas selected in each country.
Survey of street children
Due to the special problems of collecting data on children working and living on the streets (i.e. children not residing in a household), an individual questionnaire was formulated and used to assemble information on variables relating to the schooling and non-schooling activities of such children, their living and working conditions, parents, migration status and so on. Given that children working on the streets often live on their own and do not have a usual place of residence or home, they could not be represented in a sample of households.
Therefore, for the children on the streets, a "purposive" or "convenience" approach was applied and, as far as possible, enumerators who were selected and trained for the interviews were those who knew the core areas where such children were found. The children were visited in their localities in the evenings, and in some cases at night if that proved to be more convenient. In the urban core, many children tend to form groups to eat and sleep together.
The time-use approach
The methodological experiments included a "time use" module for interviewing individual children within the households and working and living on the streets. A list of economic and non-economic activities was constructed and used to identify which activities children had been engaged in during the 24 hours preceding the time of the survey and to determine how much time had been devoted to each activity.
[Ukrainian] [English] [Russian]