Brazil
Title of the survey
Relaçao Anual de Informaçoes Sociais (RAIS) (annual social information
report)
Organization responsible
Ministerio do Trabalho e da Previdência Social (MTPS), Secretaria
Nacional de Trabalho (SNT).
Periodicity of the survey
Annual
Objectives of the survey
To identify employees on the payroll who have a recognized entitlement
to a wage or salary; to assist in checking the contributions to the
security fund for years of service (FGTS); to provide information on
the formal labour market; to make available information of use for
technical studies in the social area; to assist in the compilation of
statistics and in related investigations; to act as an instrument for
taxation and for checking that persons subject to legal norms and
requirements comply with them.
Main labour topics covered by the survey
Employment, earnings and contractual wages and salaries, and
fluctuations in employment.
Reference period
For employment and contractual wages and salaries: 31 December of
each year.
For earnings: every month of the year.
Coverage of the survey
Geographical
The whole country.
Industrial
All branches of economic activity.
Establishments
Establishments of any type and of any size.
Persons
All employees.
Occupations
Some 70 occupational groups were selected as the most common at the
national level.
Concepts and definitions
Employment
An employment relationship (i.e. a job filled by an employee) is the
relationship governed by the Consolidated Labour Laws (CLT). Such a
relationship is established when a person needs paid employment and
enters a subordinate relationship to an employer, who lays down the
hours of work beforehand. The work relationship established between an
unemployed person and the trade union which recruits him is also
regarded as an employment relationship.
Among employees, the following categories are distinguished:
- employees recruited in accordance with the regulations of the CLT
for an unspecified period, including those recruited for the purpose of
acquiring work experience,
- statutory employees (officials) in the public service (both direct
and indirect), the Federal service, the service of the states,
corporations and municipalities, and officials in supervised
institutions,
- unemployed workers recruited by trade unions,
- temporary workers covered by Law 6.019 of 3.1.1974,
- directors who have no employment relationship, whose pension the
enterprise has decided to transfer to the FGTS,
- retired civil servants,
- workers regulated by the statute on rural workers, provided that
they have been employed by a legal person.
In addition, the number of proprietors and unpaid family workers
working during the reference year is ascertained, but such persons are
not included in the statistics.
Independent and temporary workers and members of courses regulated by
MTPS Decree No. 1002 of 29 September 1967 are excluded.
Earnings
These refer to gross remuneration, before deduction of taxes
or social security contributions. They include the following
components:
- wages, salaries, benefits, supplements, daily allowances, fees,
additional perquisites, overtime pay, additional pay, increments, tips,
gratuities, shares of profits, interest, percentage payments,
commissions and paid leave;
- the total value of cost-of-living allowances, when these exceed 50
per cent of wages or salaries, unless these are paid from public funds;
- gratuities and shares in profits, productivity, seniority, special
duties and responsibilities, and vacation bonuses when these exceed 20
days' pay;
- compensation to employees who retire from managerial duties and
payment to directors who have no employment relationship but who have
the right to a pension from the FGTS;
- vacations granted as compensation, additional bonuses for years of
service or for special duties or responsibilities, where these exceed 20
days' pay (as in the case above);
- total payment for the period in days, multiplied by 50 per cent of
this value if twice the ordinary rate is paid;
- payment to external accountants who have an employment
relationship, and supplementary payments for overtime, night work and
urgent or difficult work;
- days paid as compensation at the end of the contract and maternity
payments or leave granted for pregnancy announced in the month when the
contract terminates;
- the value of cash benefits such as food, transport, housing and
clothing.
These components are not listed separately in the RAIS.
Data are also collected on contractual wages and salaries
as at 31 December each year.
Wage/salary rates
Not relevant.
Hours of work
Data are collected on normal hours, i.e. on the hours worked
per week, or the number of hours of work performed per month by the
employee, not counting overtime. Normal hours are the hours laid down
by laws, regulations or collective agreements. The unit of measurement
is the month.
International recommendations
The concept of gross remuneration used in this survey is similar
to that of total earnings actually paid, which appears in the
international recommendations on wage statistics. The concept of
normal hours matches the concept included in the international
guidelines.
Classifications
Industrial
The Classifiçao Nacional de Actividades Económicas (CNAE) (national
classification of economic activities) is used, at the level of the nine
major one-digit divisions. The CNAE is linked to the International
Standard Industrial Classification of All Economic Activities (ISIC),
Rev. 2, 1968.
Occupational
70 of the most common occupational groups are used, together with
10 major occupational groups linked to the International Standard
Classification of Occupations (ISCO-68).
Others
The data are classified by sex, nationality, level of education, region,
duration of service (seniority), type of employment relationship, etc.
Sample size and design
Statistical unit
The unit of observation is the establishment; this is deemed
to be all units of an enterprise at separate locations. Thus, for
example, if the various production lines of an enterprise are located in
the same place, they are regarded as one establishment.
Survey universe / sample frame
The RAIS register covers all establishments and physical persons
who, as employers, have maintained an employment relationship at any
time during the year, independently of the organization of the
enterprise, together with establishments which have not kept up
employment relationships and trade unions which bring together
unemployed workers. These establishments are required, by Decree No.
76.906 of 1975, to make an annual declaration to the RAIS at the
beginning of the year following the employment relationship which is
being surveyed. This makes a total of 2.5 million establishments and
38 million employment relationships, covering essentially 80 per cent of
the total survey population.
The register is updated every year, using the data supplied.
Sample design
The administrative register is in practice a census of the labour
market. It covers some 2.5 million establishments and 36 million
persons employed under an employment relationship.
Field work
Data collection
The information is collected between January and April each year. The
employers obtain the forms from stationers or use magnetic tape to
transmit the data. On the back of the forms there are general
instructions with all the information needed for filling them in
properly.
Survey questionnaire
Not available.
Substitution of sampling units
Not relevant (the RAIS is an administrative register).
Data processing and editing
Processing of the data is the responsibility of the Federal data
processing service (SEPRO), a public enterprise linked to the Ministerio
de Economía, Hacienda y Planificación. The replies are obtained on the
basis of the coded form or from magnetic tape, and the data received are
processed by computer. When errors or omissions in the information are
found, they are corrected by data from the registers of the social
integration programme (PIS) and of the training programme for
public servants (PASEP) or of the general register
of contributors (CGC) and other special records.
Types of estimates
Totals of employment and monthly averages of wages and contractual wages
and salaries.
Contractual wages and salaries have always been dealt with in monthly
equivalents. Payment for piece work and other types of wages are
included in a not declared category.
Earnings are always considered on a monthly basis, i.e. as if the
workers had worked for the whole month, except for the months when they
entered and left employment; for these the necessary correction is
made. In calculating attendance, it is assumed that the employees work
throughout each month, except for the months when they entered and left
employment, in which case the proportion actually worked is taken into
account. In all relevant calculations the months in which no monthly
earnings were received are ignored.
Construction of indices
Index numbers of employment are constructed. The method is established
for consecutive two-year periods as follows:
- At the first stage, fixed base indices (1979=100) are obtained for
each sphere of interest, by means of successive linking of the indices
of change for each year with those of the previous year. The annual
indices are calculated separately for each period, by ascertaining the
difference between the total employment (T) at the end of the periods t
and t-1, which correspond to totals of identical units (n) in this area
submitted to the RAIS in both periods. Secondly, those establishments
which report to the RAIS that they have no employees on 31 December of
two years t or t - 1, are left out of the calculation of the indices.
- At the second stage, annual estimates of employment T are obtained
for each area in the period covered by the fixed base indices.
Weighting of sample results
Not relevant.
Adjustments
Non-response
Information on employees is obtained through the records of PIS/PASEP.
Other bias
With respect to earnings, for analysing the consistency of data on wages
and earnings, a double criterion of adjustment is used, up to a maximum
limit of earnings, comparing these with the variables of occupation and
level of education, the values adopted for these being the maxima
between the two criteria employed. Contractual wages and salaries with
zero declaration remain zero; the contractual wages and salaries of a
non-valid class are regarded as monthly salaries.
Use of benchmark data
Not relevant.
Seasonal variations
Not relevant.
Indicators of reliability of the estimates
Coverage of the sampling frame
The RAIS register covers essentially 80 per cent of the
total survey population.
Sampling error / sampling variance
Not relevant.
Non-response rate
Approximately ten per cent.
Non-sampling errors
Biases in the results can arise from:
- errors in information with regard to the dates of the beginning and
end of contracts of employment,
- errors in information regarding the level of education and
occupation (occupations not covered, employees not declared, etc.),
- errors in coding the economic activity or in the location of the
establishment,
- contractual wages and salaries not covered or earnings declared
after the time specified for the dates of beginning and leaving
employment.
Conformity with other sources
It is not possible to compare RAIS with other sources, since by its
nature RAIS comprises a census of the labour market, which no other
organization carries out in Brazil. The other sources mentioned are
used in order to supplement information which is provided to the RAIS
but which is ambiguous, or to provide information which is missing.
Available series
Data are presented in the RAIS Yearbook in four sections:
- number of employees on 31 December by age, level of education, size
of establishment, nature of the employment relationship, years of
service, earnings, contractual wage or salary, economic activity,
nationality and occupation;
- fluctuations in employment (entering and leaving), in accordance
with various cross-classifications;
- contractual wages and salaries as at 31 December and average
earnings for each month of the year, by various characteristics;
- establishments according to geographical and sectoral distribution
and, by size during the reference period.
History of the survey
The Annual Social Information Report (RAIS) was set up by
Decree No. 76.906 of 1975.
The estimates of the RAIS are subjected to a complete
revision every year once the results of all the most recent
tables are obtained, in order to bring the indices up to date and
set up for the RAIS new nurseries of data on employment.
Documentation
Ministério de Trabalho, Secretaria Geral Anuário RAIS (annual;
Brasilia). This appears about two years after the reference year.
Data can also be supplied on magnetic tape and cassettes, and results
which do not figure in the publications can be obtained on application
to the Ministério de Trabalho.
Confidentiality / Reliability criteria
Data which permit the responding units or employees to be identified may
not be published. This prohibition was laid down in the National
Statistics Law and the Constitution.
Other information
Data supplied to the ILO for publication
The following series are published in the Yearbook of Labour
Statistics
- paid employment in manufacturing (in Table 5B), and
- average monthly earnings of employees in non-agricultural activities
and in specific industries (in Tables 16, 17A, 18, 19 and 20).