In October 2000, the Statistics and Census Office (Dirección de Estadística y Censo), together with the International Labour Office, carried out the Child Labour Survey, in order to provide information which would allow an evaluation of the impact of children’s labour market participation, in order to determine the characteristics and conditions under which this labour market participation occurs, its possible causes, and the existence or not of exploitative relationships. The Survey provides crucial information for preparing specific policies for the population between the ages of 5 and 17 years, as well as for monitoring and evaluation of programmes being carried out by different social agencies attempting to eradicate the worst forms of child labour.
The Child Labour Survey was carried out guided by the following objectives:
- Ascertain demographic and socio-economic characteristics of the general population and especially the child population
- Provide information that will allow studies of the magnitude, distribution, characteristics, consequences and causes of child labour
- Ascertain characteristics of economic sectors where minors are working
- Ascertain injuries sustained by the employed population
- Ascertain safety mechanisms available to the employed population
- Ascertain parents’ perceptions and those of children regarding child labour
- Provide a database on child labour that permits formulation of policies and programmes based on reality
- Provide information allowing cross-country comparisons
The main topics investigated in the suvey include the following:
- Housing characteristics
- Household income
- Household expenses
- Population size
- Socio-demographic characteristics
- Educational level
- Reasons for dropping out of school
- Current employment
- Economic sector
- Occupational category
- Generation of revenue, pay and other benefits
- Occupational injuries
- Benefits received from employer
- Parents perceptions of children’s jobs
- Children’s perceptions of work
- Participation in household chores
The study universe is the population 5 to 17 years of age residing in private occupied dwellings throughout the country. The interim results from the Population and Housing Census of May 2000 found a preliminary total population of 2,815,644 persons for the country; of these, 766,903 constitute the population from 5 to 17 years of age (see Table 3 in the report provided as external resources),which implies a percentage relationship of 27.2%. This population is divided 56.9% urban and 43.1% rural. Furthermore, the non-indigenous universe contains a 5 to 17 year old population of 693,704 persons and the indigenous one has 73,199, which represent 90.5% and 9.5% of the study population respectively. Private occupied dwellings numbered 667,284 units at the national level,with urban areas representing 64.3% and rural areas 35.7%. In non-indigenous areas, private occupied dwellings numbered 638,565 units, while indigenous areas had 28,719, for a percentage relationship of 95.7% and 4.3%, respectively. The average number of persons aged 5 to 17 years per private occupied dwelling in the country was 1.15 persons per dwelling, 1.02 in urban areas and 1.39 in rural areas, while for the non-indigenous universe this was 1.09 and for the indigenous 2.54.
Producers and sponsors
Statistics and Census Office (Dirección de Estadística y Censo)
Contraloría General de la República de Panamá
International Programme for Elimination of Child Labour
Sampling Frame: With preliminary data from the Population and Housing Census as a reference, as well as the full census organisation and maps from May 2000, the sampling frame was made up of the enumeration area units where population aged 5 to 17 years was recorded.
Sampling units: The sampling units constitute the sample selection unit. In this case, the Primary Sampling Unit is the census segment.
Study domains: The study domains were identified,with a view to type of study and user requirements regarding utility and utilization of information. The country's main province, Panama, was subdivided into the following study domains: Panama and San Miguelito districts and rest of Panama province. The indigenous study domain is integrated at the national level by each one of the legally established comarcas and the indigenous communities outside the comarcas that carry out their activities according to their socio-cultural behaviour patterns.
Stratification: Study universe stratification is based on geographic criteria in accordance with the country's political-administrative coding and takes into consideration the division between urban and rural areas. Stratification by socio-economic variables was not possible, since the census information was not yet ready, as complete processing was expected for March 2001. It is important to note that in Panama implicit stratification has been used in several studies such as, for example, the Quality of Life Survey, the Income and Expense Survey, the Labour Force Survey, etc., obtaining adequate results with regard to the existing socio-economic structure, supported primarily by the particularities occurring in geographic distribution of the country's population.
Sample size: Sample size computation was carried out using the mathematical model for simple random sampling.The critical variable for obtaining sample size is represented by proportion of population aged 5 to 17 years. For Darién and Los Santos provinces and Panama and San Miguelito districts sample size calculations were independent for urban and rural areas, due primordially to the fact that sizes obtained by proportional distribution in some areas were very small, leading to a sampling error much larger than desired.
Sample selection: Probabilistic cluster sampling was applied. Sample design responds to a one-stage design, selecting primary sampling units (census segments) by systematic selection with probability proportional to size. The statistical inference process for the nonindigenous sample selected was carried out by area, by applying the ratio method, using as the exogenous variable demographic projections of the population 5 to 17 years of age.
Pages 6-8 of the study report (available as external resources) provide detailed explanation, formulae and tables on sample design and implementation.
Dates of Data Collection
Data Collection Mode
On the average, each interviewer was responsible for three segments per week. Forty-four supervisors were assigned, each one of whom supervised an average of three individuals. This guaranteed direct fieldwork supervision and that corresponding verifications could be carried out. The investigation lasted one month; during this time 1,622 segments were investigated.Forty-two percent of these were in urban areas, 54% in rural areas and 4% in indigenous areas. These segments were visited weekly, and as was indicated above, customary residents of dwellings were interviewed.The questionnaires were revised and verified by supervisors and forwarded to the central office one week after the activity was completed.
Table 2 in the survey report (provided as external resources) indicates the number of supervisors and interviewers recruited and required for each province.
Data Collection Notes
The Human Resources Office of the Comptroller General handled personnel recruitment. Once recruited, individual interviews were carried out with each candidate, who filled the following profile:
- Completed secondary education in sciences, letters or commerce, preferably with university studies in majors related to social sciences or similar studies.
- Ease of expression.
- Preferably with some experience in studies of this type.
- Available for work in the area assigned, whether or not s/he lives nearby and whether it is urban or rural.
- Have a sense of responsibility.
- Outgoing personality.
- Good physical condition.
- Legible handwriting.
- Available for work any day of the week, including Saturday and Sunday.
In order to provide a small scale reproduction of each aspect related to implementing the Survey, a pre-test was arranged for August 16th through 25th, 2000, where each instrument to be used in the Child Labour Survey was tested and refined. Details on the pretest are available on pages 3-4 (Section 2.B.2) of the survey report provided as external resources.
Child Labour Survey Fieldwork
During October 2000, the Statistics and Census Office carried out fieldwork for the Child Labour Survey. To accomplish this activity, it was necessary to carry out a number of tasks both before and after the fieldwork itself. The general objective of the Survey was to measure child labour. More specifically, there was interest in measuring socio-economic characteristics of dwellings with population between 5 and 17 years of age, ascertaining housing conditions of the child population entering the labour force, measuring the conditions under which child labour occurs, and obtaining information on occupational hazards and lesions occurring to the
population of boy, girl and adolescent workers.
Training in all of provinces except Bocas del Toro and Darién was carried out between October 5th and 10th, 2000. In these two provinces it was held between the 8th and 13th of October 2000. Training lasted six days and required personnel were trained, along with one reserve. The number of staff trained by province was as following: in Bocas del Toro 14, in Coclé 16, Colón 15, Chiriquí 19, Herrera 19, Los Santos 22, Veraguas 17, Darién 10 and Panama 45. Additionally, the team of instructors handled supervisor training at the national level in Panama province on October 2nd and 3rd.
Fieldwork was carried out with personal interviews throughout the country, in those dwellings where a population aged 5 to 17 years had been detected, regardless of the fact whether this population was working or not. The interviewers visited segments situated in urban, rural, indigenous and difficult to reach areas. Survey coordinators were named on a province basis; they were responsible for coordinating administrative aspects of the survey. While the survey was being carried out, there were periodic fieldwork and general progress evaluation visits in each of the provinces.
In order to have more effective control over the survey, a segment route form was used, avoiding all possible omissions by the interviewer; i.e., the interviewer visited all dwellings in the segment and carried out interviews in those dwellings where persons aged 5 to 17 years old resided. When a dwelling was visited and the population of interest was not found, dwelling information was recorded on the form designed for that purpose and the situation of the dwelling was recorded next to it.
The questionnaire is divided into the following sections:
I. Location of the dwelling
II. Information on the dwelling
III. Household income and expenses
IV. List of occupants
V. General Characteristics
VI. Sociodemographic Characteristics
VII. Educational Characteristics
VIII. Economic Characteristics
IX. Job related injuries or diseases
X. Parent perceptions of persons aged 5-12 years who are employed
XI. Perceptions of persons aged 5-12 years who are employed.
Spanish and English versions of the questionnaire are provided as external resources.
In order to facilitate development of the different systems, the following documentation was available: list of interviewers and supervisors by province and code; recode list for conditions of employment; list of validations and inconsistencies; interviewer's manual; final questionnaire; file of segments covered by the Survey and a file of indigenous segments. Six systems were developed for the Child Labour Survey, to wit: Data Entry System, Coverage Control System, Recode System, Tabulation System, Expansion Factor System and Data Dictionary System. The Data Entry System was developed using Visual FoxPro (Release 5.0). This system is divided into 4 Sub-Systems: Addition, Query, Modification and Elimination.Validations and inconsistency correction were carried out on-line. This means that the System would not allow data entry personnel to continue if they had not made the due corrections. For greater security and data integrity, once the data entry period was over, batch verifications were carried out, using the same package mentioned above.
The Data Entry System was developed under Client-Server architecture. This means that the executable system was server-based, along with its different components, including the 5 databases which were receiving information from questionnaire contents. The Client was a data entry person using a PC as a terminal to access the Server. Every afternoon a backup was carried out from the server to the PC of the Child Labour Survey Information Systems Administrator. Data entry personnel for this Child Labour Survey were chosen from a larger group of data entry personnel from the Household Survey, in addition to having had fieldwork experience. Data entry personnel were limited to consulting specialised personnel or the Programmer-Analyst for the different Systems for this Child Labour Survey. The data were collected in Spanish. For the Statistics and Census Office, any information requested from a citizen is confidential material. Thus, for data entry personnel, the questionnaire and everything related to it was sensitive information. There were controls to determine which data entry person was entering which folder and on which date. The questionnaires could not be removed from the work area.
Each data entry person (of the 8 selected) removed a folder from the shelves, after writing it down in the control list. Each folder was made up of 4 to 6 segments, depending on their size. Data entry began on March 28 and ended on June 12 of the same year, 2001. During the first month, data entry worked in two shifts, from 7:00 a.m. to 3:00 p.m. and from 3:00 p.m. to 10:00 p.m. Later, a single shift was put into effect, from 8:30 a.m. to 4:30 p.m. Initial verification of data entry for a questionnaire was the responsibility of a database called Coverage, which verified that a particular segment was valid (Province + District + Corregimiento +
Segment). If the foregoing was correct, data entry could continue for Questionnaire + Household + Person Number (this last number was excluded for the Dwelling database). These data, together with that from Coverage, were verified in the other 5 databases to avoid duplicating keys. Data for a total of 9,261 questionnaires were entered.
It should be mentioned that the data processing directives followed are those established by the National Computing Office (Dirección Nacional de Informática), which are revised each year. The Population and Housing Census, which is carried out every ten years acts as a conceptual framework for all data processing within the Statistics and Census Office.
Details on other processing are provided on page 10 of the survey report available as external resources.
Estimates of Sampling Error
Sampling error levels by area for each province in the study vary between 3% and 5%, with the exception of Darién province, which due to its urban population composition has a theoretical tolerance level of 8%, while the rural area is below 5%. Computation of the indigenous sample size was carried out at the national level, with a confidence level of 95% and a sampling error of 4%. Sample size was adjusted taking into consideration the four weeks during which the Survey would be carried out, which resulted in a sample larger than that calculated.
Statistics and Census Office (Dirección de Estadística y Censo)
Use of the dataset must be acknowledged using a citation which would include:
- the Identification of the Primary Investigator
- the title of the survey (including country, acronym and year of implementation)
- the survey reference number
- the source and date of download
Statistics and Census Office (Dirección de Estadística y Censo)
The user of the data acknowledges that the original collector of the data, the authorized distributor of the data, and the relevant funding agency bear no responsibility for use of the data or for interpretations or inferences based upon such uses.