Trinidad and Tobago

1.Title of the survey:

Continuous Sample Survey of Population (CSSP).

2.Organization responsible for the survey:

Central Statistical Office, Office of the Prime Minister.

3.Coverage of the survey:

(a) Geographical:

The whole country.

(b) Persons covered:

All persons who are normally residents in private households, as well as some of the persons staying in institutions and for whom information is collected at the private household to which they belong. Also included are private households located within institutions.

Persons are considered as normally residents at a given address if they normally sleep most nights of the week at this address, or if they are found at this address and although they are residents of Trinidad and Tobago, have no other usual place of residence. Included are boarders and lodgers, as well as domestic servants.

Persons in institutions such as hotels, large boarding houses, hostels, barracks, etc. (Group A) are included in the survey and information obtained for them from the household where they normally reside, regardless of the length of time they have been at these institutions.

Inmates of institutions such as hospitals, nursing homes, prisons, leprosaria, etc. (Group B) are included in the survey and information obtained for them from the household where they normally reside, if they have been living in such institutions for less than six months.

Excluded from the survey are inmates of Group B institutions who have been in such an institution for six months or more, as well as those few persons in Group A institutions who have no other place of residence in Trinidad and Tobago.

The economically active population refers to persons aged 15 years and over.

4.Periodicity of the survey:

The survey is bi-annual and carried out in 12 periods within each six-month round (January-June and July-December).

5.Reference period:

The calendar week immediately preceding the week of enumeration (referred to as the survey week).

6.Topics covered:

The CSSP provides information on employment, unemployment, hours of work, duration of unemployment, industry, occupation, status in employment and level of education.

7.Concepts and definitions:

(a) Employment:

The employed comprise:
  1. "all persons who worked for pay for any length of time during the survey week;
  2. persons who were temporarily absent from work because of vacation, illness, industrial dispute or some similar cause, but who had jobs to which they would return at the end of that period of absence; (seasonal workers out of work are not included in this category); and
  3. persons who worked without pay on a family farm or business or as a learner, irrespective of the number of hours worked."

This definition includes:

  1. full- and part-time workers seeking other work during the reference week;
  2. full- and part-time students working full- or part-time;
  3. persons who performed some work for pay or profit during the reference week while being subject to compulsory schooling; or retired and receiving a pension; or registered as jobseekers at an employment office or receiving unemployment benefits;
  4. paid and unpaid apprentices and trainees;
  5. participants in employment promotion schemes;
  6. private domestic servants;
  7. members of producer's co-operatives;
  8. members of the armed forces (when they are covered by the survey - see under Coverage) and persons engaged in civilian service equivalent to military service.

Work for pay covers all activities for which an individual receives remuneration, including work in one's own business or farm if run for profit; odd jobs undertaken for pay; and subsidiary activities. By this definition, are included as working, for example, housewives who during the particular survey week did some dressmaking or hairdressing for pay even if they do not usually work for pay.

Excluded from the employed and considered as inactive are persons engaged in own housework; and those doing unpaid community or social work.

(b) Underemployment:

The information is not available.

(c) Unemployment:

The unemployed are "all persons who looked for work during the three-month period preceding the enumeration and who, at the time of enumeration, did not have a job but still wanted work."

The unemployed are subdivided into:

  1. persons without jobs and seeking work during the survey week, who are persons who actually made an effort to obtain a job, or indicated that they wanted a job but did not look for work for one of the following reasons:
    1. they were on lay-off without pay;
    2. they were awaiting the results of previous applications for employment;
    3. they were temporarily ill and could not look for work in that week;
    4. they wanted a job but did not know where to get suitable employment.

  2. persons without jobs who looked for work at some time during the three-month period preceding enumeration but not during the survey week, and are still available for work.

Persons not at work nor seeking work because they were temporarily ill, or they were awaiting results of previous applications for employment, or they knew of no suitable vacancy, or they had some other reason for not looking for work, which was to be specified, are classified according to the length of time since they last looked for work. Among those persons, those who actually sought work at some time during the three-month period preceding enumeration are included as unemployed.

Also classified as unemployed are full- and part-time students seeking full- or part-time work.

Persons are classified as seeking jobs as employees if, during the three-month period preceding enumeration, they applied in person or in writing to an employer, or went to a labour office seeking work, or made some other effort to get a job. Persons are classified as seeking self-employment if they were making plans to work on their own, or trying to make arrangements to set up a business, or begin some economic activity as a self-employed person.

(d) Hours of work:

All persons who had a job in the survey week are classified according to the number of hours they actually worked. Periods of absence from work because of vacation, leave, illness, etc. are excluded. The information is collected in ranges of hours, i.e. from 1 to 8 hours; 9 to 16; 17 to 24; 25 to 32; 33 to 40; 41 hours and more. The number of hours relates to completed hours (e.g. 8 hours and 50 minutes are classified as 8 hours).

(e) Informal sector:

The information is not available.

(f) Usual activity:

The information is not available.

8.Classifications used:

Employed persons and unemployed persons with previous work experience are classified by industry, occupation and status in employment. All persons covered by the survey are classified according to their level of education.

Employed persons with more than one job are classified according to their main job; unemployed persons with previous experience are classified according to the last job they held.

(a) Industry:

The classification used is compatible with the first revision of the International Standard Industrial Classification of all Economic Activities (ISIC-1958) at the major division level.

(b) Occupation:

The classification used is compatible with the International Standard Classification of Occupations (ISCO-1968) at the major group level, with some exceptions.

(c) Status in employment:

The following status groups (named Types of Worker in national publications) are used:
  1. Employee, subdivided into government employee and other,
  2. Unpaid worker (i.e. unpaid family worker or learner),
  3. Employer,
  4. Own-account worker.

(d) Level of education/qualifications:

The information is not available.

9.Sample size and design:

(a) The sample frame:

It consists of the listings of enumeration districts, compiled from the decennial population census and updated at each period of listing in order to maintain precision of the estimates. The sample design is modified after each census to reflect changing situations.

Based on the 1970 Census of Population and Housing, the country was divided into nine domains of study, and enumeration districts (EDs) were defined to fit within specific natural boundaries and to be of a population of 150-200 households in urban and built-up areas and approximately 70-100 households in rural areas.

EDs were assembled into homogeneous strata within each domain, on the basis of stratifying variables such as the proportion of male workers who were classified as:

  1. Professional and technical workers,
  2. Administrative, executive and managerial workers,
  3. Agricultural workers, and
  4. Male workers engaged in mining and quarrying and refining of petroleum products.

For purposes of facilitating the selection of the cluster of households to be enumerated within an ED, each ED was given a measure of size that was a multiple of approximately five households.

Based on the results of the 1970 Census of Population and Housing, approximately 48,600 clusters were derived. These were allocated among the nine domains of study approximately proportionally to the size of the population in each domain.

Utilizing data from the 1980 Census of Population and Housing, the sample design was modified to incorporate changes in the spatial distribution of the population as well as socio-economic conditions. The domains of study were increased from 9 to 14, with independent sampling selections taking place in each domain of study. (No additional information is available at present.)

(b) The sample:

The sample design is a two-stage stratified cluster sample consisting of EDs (primary units) at the first stage and non-compact clusters of households (ultimate sampling units) at the second stage.

The first stage involved the selection of a "grand sample" of clusters of households within PSUs, which consisted of nine replicates chosen with an overall sampling fraction of 1/25. Systematic selection of PSUs was undertaken with probability proportionate to size.

From this "grand sample" one-third of the nine replicates constitute the second stage sample, which is used to obtain survey data on the basis of a round of six-month duration.

From each sample ED, a non-compact cluster of households is selected systematically with a random start, and an interval equal to the number of clusters allocated to the ED. These selected clusters make up the total of households to be interviewed for a particular period of enumeration within a round.

The sample consists of approximately 3,200 households per round, which corresponds to about 1.5 per cent of the total population.

In order to maintain precise sampling probabilities, the measure of size of the sampled EDs is updated prior to sample selection.

(c) Rotation:

Two thirds of the sample are repeated between successive rounds, with 1/3 being rotated out.

10.Field work:

(a) Data collection:

Data are collected by personal interview. A permanent survey organisation is responsible for data collection.

The sampling exercise consists of 12 periods within a six-month round. In each period, the first week is devoted to listing the households within the selected EDs and the second week is assigned for enumeration of the selected households. Enumeration is therefore carried out in the second and fourth week of each month.

(b) Substitution of ultimate sampling units:

In case of absence or total non-response, no substitution is made.

11.Quality controls:

Not available.

12.Weighting the sample:

The estimated total population at the middle of each round is used to blow up the sample estimates, i.e. the population as at 31st March is used for the rounds covering the first six months of the year, and the population as at 30th September for the rounds covering the months of July to December.

13.Sampling errors:

Standard error of estimates
(Round 46 - July to December 1986)
TotalMalesFemales
Employment (size of est.) 387,900260,900127,000
Standard error 5,1183,4263,377
Unemployment (size of est.) 84,20052,60031,600
Standard error 3,1402,4421,953

14.Adjustments:

(a) Population not covered:

Not available.

(b) Under/overcoverage:

Not available.

(c) Non-response:

The non-response rate is approximately 10.8 per cent. No adjustment is made for non-response.

15.Seasonal adjustment:

No adjustment is made for seasonal variations.

16.Non-sampling errors:

Not available.

17.History of the survey:

Information on several characteristics of the labour force has been collected through the Continuous Sample Survey of Population (CSSP) since May 1963. There was however a break in continuity in 1972 when the survey was used primarily for gathering information on households income and expenditure; in 1976 when preparing the Fertility Survey of 1977; and also in 1980 when the Population and Housing Census was conducted.

Rounds 1 and 2 only covered persons in Trinidad. Since round 3 (i.e. from August 1964), both Trinidad and Tobago are covered by the survey.

Data from 1980 onwards can be regarded as a new series, since population estimates have been revised in the light of the 1980 Population Census and this causes labour force estimates to appear to be lower than the 1979 estimates.

In addition, labour force data from round 42 (July to December 1984) have been revised based on the availability of more accurate information on births and deaths from the Vital Statistics Division of the Central Statistical Office.

18.Documentation:

The results of the CSSP Labour Force are published in two series:

  1. Central Statistical Office: "CSSP Labour Force, L.F.1" (bi-annual) (Port of Spain); presents the labour force and its major components as well as the total non-institutional population aged 15 years and over by sex, age group, industry, occupation and type of worker (status in employment);

  2. idem: "CSSP Labour Force, L.F.2" (bi-annual) (Port of Spain); includes tables by sex and administrative area, educational attainment, number of hours worked, and for the unemployed, the interval of time since they last worked.

Each issue is published about 18 months after the survey round it refers to and contains estimates for the previous three rounds.