LABOR MARKET - EMPLOYMENT

Contact Person

Cimar Azeredo Pereira

Organization

Phone Number

Fax Number

E-mail

cimar@ibge.gov.br

Coverage characteristics

The disseminated data are the number of employed persons aged 10 and over, in the reference week, by sex and age groups, by activity sector of the main job and its distribution by status in employment, besides the distribution of the employed persons by years of schooling.

Employed persons during the reference week are all persons who did any work for at least one hour, as paid or as unpaid workers, all those who were not working but had jobs or businesses from which they were temporarily absent because of vacation, strike, temporary suspension of labor contract, leave paid by the employer, bad weather, child-care problems, paid leave given by Social Security for no longer than 24 months, paid leave given by the employer for maternity, illness, accident for no longer than three months and voluntary absence or another reason for no longer than thirty days

The data are obtained from "Pesquisa Mensal de Emprego - PME" [Monthly Employment Survey (PME)], which is a household sample survey. Since 1980, PME has been carried out in six metropolitan areas (Recife, Salvador, Belo Horizonte, Rio de Janeiro, São Paulo and Porto Alegre), which together represent around 25% of the total population of the country and 26% of the population aged 10 and over. The results are produced for each of the six metropolitan areas and for the set of the six metropolitan areas surveyed since 1980.

In October, 2001, PME went through a methodological revision process aiming to update its thematic coverage and to adjust to the latest international recommendations. The results of the revised survey substituted for the previous series in December 2002.

The survey sample involves around 42,000 households every month.

The survey indicators have not been seasonally adjusted yet.

(For further details, refer to the Summary Methodology).

Periodicity

Timeliness

Advance dissemination of release calendar

Simultaneous release to all interested parties

The results are released in a press conference in Portuguese. The press release is made available to journalists at the start of the press conference. The "Coordenação Geral de Comunicação Social do IBGE" [General Coordination of Social Communication of the IBGE] is in charge of the dissemination of such monthly press release.

Also at the start of the press conference, the indicators and the commentary of the monthly report are made available on the site http://www.ibge.gov.br on the page "Indicadores" [Indicators] under the heading "Trabalho e Rendimento: Pesquisa Mensal de Emprego - PME" [Labor and Income: Monthly Employment Survey (PME), New Methodology] and in the IBGE automatic recuperation system (SIDRA) at:
http://www.sidra.ibge.gov.br/bda/pesquisas/pme/default.asp

Dissemination of terms and conditions under which official statistics are produced, including those relating to confidentiality of individual responses

1. Law 5 534 of November 14, 1968 (Law on mandate to give information for statistical purposes);
2. Law 5 878 of May 11, 1973 (Law that rules Brazilian Institute of Geography and Statistics - IBGE);
3. Decree 73 177 of November 20, 1973 that enacts these laws;
4. Law 6 183 of December 11, 1974 (Law on statistical and cartographical national systems);
5. Decree 74 084 of May 20, 1974 that approves the general plan of statistical and geographical information; and
6. Decree 3 272 of July 13, 2003 that approves the organic law of IBGE.

These laws and decrees, among other aspects, establish the mandatory character of providing information to IBGE and the assurance of confidentiality of individual responses, so that the information can only be used for statistical purposes.

These laws are available to the public in Portuguese at the Internet page of the Presidency of the Republic, under search of decrees and laws:
http://www.planalto.gov.br.

According to these laws, dissemination of data by IBGE is mandatory and free of charge.

Identification of internal government access to data before release

Directive 355 of November 5, 2007, issued by the Ministry of Planning, Budget and Management, establishes that data are provided to the authorities from the preceding list at 7:00 a.m. on the day of release.

The data are released for the press and disseminated on IBGE´s Internet website at 9:00 a.m.

The Directive and the relation of authorities of the precedence list are available on the website (http://www.ibge.gov.br).

Identification of ministerial commentary on the occasion of statistical releases

Provision of information about revision and advance notice of major changes in methodology

The methodological revisions are published. Statistics are periodically revised based on changes in the resident population projections used for sample data expansion, as a result of new information from the Demographic Census.

The Monthly Employment Survey - PME has been furnishing indicators for the conjuncture evaluation of labor market since 1980. The first complete methodological revision was applied in 1982. Adjustments restricted only to the sample plan were carried out in 1988 and 1993.

A new revision was carried out in 2001. The necessity of updating the investigation content and of modernizing the operational procedures led to a complete revision of the survey, which was implemented in October 2001 and produced official results in December 2002. Prior to the implementation, IBGE published on its website the aim of this review, the date it would be implemented, as well as promoted workshops with journalists and experts in the area to inform the major changes.

For further details, refer to the Summary Methodology.

Dissemination of documentation on methodology and sources used in preparing statistics

The methodology, including concepts and definitions, of the reformulated survey is presented on the publication “Série de Relatórios Metodológicos - nº 23 - Pesquisa Mensal de Emprego 2002”, available at IBGE bookstores and Internet: http://www.ibge.gov.br/home/estatistica/indicadores/trabalhoerendimento/pme_nova/default.shtm

The publication "Estatísticas Básicas – Séries Retrospectivas nº 10" presents an evaluation of the impact caused by the introduced changes and results of the reformulation, comprising data on education level, employment, status in employment, employees with formal contacts, worked hours, monthly wage, underemployment and job search of the active age population, among other aspects, as well as retrospective series of selected indicators.

(For further details, refer to the Methodological Summary.)

Dissemination of component detail, reconcilliations with related data, and statistical frameworks that support statistical cross-checks and provide assurance of reasonableness

Employment indicators, by metropolitan area and by the set, are published in the monthly report for the month of the survey, for the previous month and also for the same month of the previous year. This monthly report is available at the "Coordenação de Comunicação Social do IBGE" [Social Communication Coordination of the IBGE] and on the Internet: http://www.ibge.gov.br

In the monthly report the following indicators, for the employed persons aged 10 and over, are presented:

• Employed population
• Level of employment
• Employment rate
• Employment population by gender
• Employed population by age group
• Employed population by years of schooling
• Employed population by status in employment
• Employed population by activity sector
• Employed population by usual hours of work in all jobs
• Employed population by effective hours of work in all jobs
• Employed population by size of enterprise
• Employed population by length of permanence in the main job
• Employed population seeking work in the 30-day period - Total and Distribution
• Employed population contributing to social security in any job
• Employed population by kind of work contract Time - related underemployed - Total and Distribution
• Employed population whose Wage/Hours of work are below the Minimum Wage/Hour - Total and Distribution

The same data is presented on the page “Indicators” at www.ibge.gov.br and it is also possible to access the time series, since January, 1991, of indicators for the population aged 15 and over, related to the PME series under the previous methodology, such as:

• Economically active population;
• Employed population;
• Unemployed population;
• Average nominal income of the primary job effectively earned on the reference month by the employed population;
• Activity rate;
• Open unemployment rate (in the reference week and in the reference period of 30 days) and
• Average time of job search.

All this data is available on diskette and CD-ROM.

Notes

Telephone:  55 21 2142-4651
Fax:            55 21 2142-0941
e-mail: comunica@ibge.gov.br or on Internet at:

http://www.ibge.gov.br/

The document is available in Portuguese free of charge.

Telephone number: 0800 721 8181

http://www.ibge.gov.br on the site "Indicadores" [Economic Indicators] under the heading "Trabalho e Rendimento" [Labor and Income]. Available free of charge.

PME – Monthly Employment Survey available on the Internet at:
http://www.ibge.gov.br/home/estatistica/indicadores/trabalhoerendimento/pme_nova/default.shtm

IBGE system of automatic recuperation (SIDRA) at:
http://www.sidra.ibge.gov.br/bda/pesquisas/pme/default.asp

Available free of charge.

Telephone number: 0800 21 8181

Telephone number: 0800 21 8181

The Montly Employment Survey - PME aims to produce monthly indicators on labor force which enable the evaluation of the fluctuations and trends, on medium and long term bases, of labor market, on their coverage areas. It is a continuous household survey used to give an agile indicator of the effects of the economic conjuncture on the labor market, besides meeting other important needs for the socioeconomic planning of the country. At the present time, PME comprises the metropolitan areas of Recife, Salvador, Belo Horizonte, Rio de Janeiro, São Paulo and Porto Alegre. Approximately 42,000 housing units are eligible for interview every month.

The results are obtained for each of the metropolitan areas comprised by PME and for the set of six areas surveyed since 1980.

PME went through a complete methodological revision in 1982 and, in 1988 and 1993, through adjustments restricted to the sample plan. Its methodology was completely revised in 2001.

Implemented in 2001, the revision of PME enabled a more comprehensive collection of labor characteristics and the ways labor force inserted in the productive system, thus providing more information for the study of labor market and for the formulation and attendance to political policies. In order to measure the effects of the introduced changes, both surveys (old and new ones) were carried out simultaneously for 14 months, so that the results derived from the changes could be analyzed and explained.

The new survey started to produce official indicators in December 2002.

Summarizing, the main objectives of PME revision were:

1. the implementation of some conceptual changes in the subject work, in accordance with international recommendations;
2. a more comprehensive investigation which enables a better knowledge of the economically active population and of the population available for the labor market;
3. adjustment in the sample rotation process providing better conditions for the longitudinal attendance of the results;
4. turning the questions into a more operational level in order to collect information so as to improve the measurement of the phenomena;
5. the introduction of the electronic collector to accomplish the data collection operations so as to improve the operational system of the survey and to reduce the duration of data processing.

The survey methodology tries to follow the International Labor Organization - ILO recommendations. The main concepts and definitions related to the labor theme surveyed in PME are:

A) Reference Periods

The reference periods adopted are:

• Reference Week: It is the week, from Sunday to Saturday, previous to the week defined as the interview week for the housing unit.
• Reference Date: It is the date of the last day of the reference week.
• Reference period of 30 days: It is the period of 30 days which ends on the last day of the reference week.
• Reference period of 365 days: It is the period of 365 days which ends on the last day of the reference week.
• Reference month: It is the previous month which contains the four reference weeks that compose the month of the survey.

B) Employment

Employment in economic activity is considered as the exercise of :

1. paid employment in cash, products, goods or benefits (housing, food, clothing, training, etc) in the production of goods and services for at least one hour during the reference week;
2. paid employment paid in cash or benefits (housing, food, clothing, etc) in the household service for at least one hour during the reference week ; or
3. unpaid work in the production of goods and services, supporting the economic activity of a member of the household, for at least one hour during the reference week.

C) Economically active population during the reference week

The economically active population during the reference week comprises employed and unemployed persons aged 10 or over during the reference week, as follows:

D) Employed persons during the reference week

The employed persons during the reference week comprise the persons aged 10 or over who performed paid or unpaid work for at least one complete hour during the reference week or who had a job but were temporarily not at work during this week.

Since 2001, it has been considered as employed but temporarily away from paid work, the person who did not work for at least one complete hour during the reference week due to vacation, strike, temporary suspension of formal contract, leave paid by the employer, bad weather conditions or other occasional factors. Similarly, it has been considered the person who, on the reference date, was away due to leave paid by social security for a period not longer than twenty-four months; from own business due to pregnancy, illness or accident, not on leave by social security in a period not over three months; due to voluntary absence or other reason, for a period not over thirty days.According to the previous methodology, every person who had a job during the reference period but did not work was regarded as an employed person.

E) Unemployed population during the reference week

The unemployed population during the reference week comprises the persons aged 10 or over without a job during the reference week, but who were available to get a job this week and took effective steps to do that in the reference period of 30 days, without having any job after quitting the last job they had during this period.

Seeking work is regarded as taking effective steps to get a job that is, contacting to employers; taking or applying for contests; contacting job agencies, unions or similar institutions; answering or placing job advertisements; seeking assistance of relatives or friends in order to get a job; taking steps to establish own business, among others.

When an effective step to take a job is identified, we ask how long ago the last step was taken and the total duration of job seeking. (The duration of time elapsed between the last effective step to take a job and the continuous job search must be investigated in case the last step taken to get a job is identified.)

The duration of job search is regarded as the continuous period of making arrangements to take up a job. It is investigated the time the person had been taking measures to get a job; if there is an interruption of more than two consecutive weeks it should be counted up to the date of the last measure on the reference period of 365 days. The beginning of the counting is not limited by the reference period of 365 days. That is, the person may have begun to take measures to get a job before this period and all the uninterrupted time of measures must be counted. On the counting of time that the person took measures to get a job, besides having not taking measures for more than two weeks, it is also considered as interruption the period when the person carried out some job. The counting can be re-started when the person is employed, provided that he/she had resumed to take measures to get another job.

F) Population not economically active during the reference week

The population not economically active during the reference week comprises all persons aged 10 or over who were neither classified as employed nor as unemployed during the reference week.

G) Main job

For the person who had more than one job, that was attached to more than one employment, the main job during the reference week is defined as the one in which he/she effectively worked the greatest number of hours during the reference week. If the number of hours effectively worked during the reference week is the same, the main job is defined as the one in which the person spent more time counted up to the end of the reference week. In case the latter also coincides, the main job is defined as the one that usually provided the greatest monthly income.

H) Status in employment

Status in employment is known as the work relationship between the worker and the enterprise where he/she works. The employed population is classified by status in employment as:

• Employee - person who worked for an employer (individual or legal entity), carrying out a working day and getting payment for that in cash, goods, products or benefits (housing, food, clothing, training, etc). The person in the Armed Forces, the clergy (priest, minister, rabbi and others) and, also, the apprentice or trainee who received only training as payment are included in this category.
  An employee is also classified as:
  • Domestic worker - person who is exclusively engaged by households to render domestic services, paid in cash or benefits, for one or more households units.
  • Unpaid family worker who helps an employee member of the household - person who worked for at least one hour per week, helping member of the household, with whom the employer had a job contract and that received the payment for the work of group of members of the household he/she organized, directed or was responsible for.
    
• Own-account worker - person who had his/her own business, alone or in partnership, without any employees and with or without help from unpaid worker who is also a member of the household.
• Employer - person who had his own business, with at least one employee, with or without the help from unpaid worker who is also a member of the household.
• Unpaid family worker - person who performed unpaid work on enterprise of member of the household who was own-account worker or employer.

I) Employment – related Income

• For employees, the income from paid employment consists of all payments in cash, products or goods. The value of the payment received in benefits (kind and services) - giving or paying for, directly by the employer, like housing, clothing, food, transportation, training or education - is not computed.
  • Gross income in cash - gross income from work in cash, composed by one or several categories (salary, wages, gratuities, cost-of-living allowances, family allowances, yearly basis or five-year basis bonuses, bonuses, premium pay for overtime, benefits paid in cash and others without deducting the employees’ direct fees, transfers and others obligations (such as: social security contribution , income tax, pension schemes, labor union contribution, security and health plans, etc).
  • Gross income in products or goods - gross work income in products or goods of the activity which comprises agriculture, livestock breeding, hunting, forestry, extractive agriculture, fishing and aquiculture, computed by its value in cash, excluding the amount destined for the household consumption.
    
• For the own-account workers and employers, the withdraw of cash or goods is considered:
  • Withdrawal in cash - fixed withdrawal or as a profit percentage of the enterprise, without excluding personal payments (for instance, social security contribution and income tax of the person ) or when the business is not organized in such a way the income in cash or work done is directly identified as the difference between the income and the expenditure (payment of employees, raw materials, electricity, telephone, equipment and other investments) of the enterprise.
  • Withdrawal in products or goods - withdrawal in products or goods, from the activity which comprises agriculture, livestock breeding , hunting, forestry, extractive agriculture, fishing and aquiculture, computed by its value in cash as the difference between the value of the products and goods destined for the market and the necessary expenses to its production, excluding the amount destined for the household consumption.

Since the latest revision took place, two kinds of earnings have been investigated for all paid workers: usually received earnings and effectively received earnings.

Monthly income usually earned from employment

The monthly income usually earned from employment is regarded as the one the person usually earned in a complete month of work.

In case of fixed cash remuneration, the monthly income is regarded as how much the person usually earned, in terms of the reference month in which the reference week was. In the event of a variable cash remuneration, the monthly income is the average that the person earned, in terms of the reference month. When the remuneration in cash varied on account of the period or season of the year, the monthly income is regarded as the one the person usually earned in the seasonal period in which the reference week was.

In relation to the remuneration in products or goods, from the activity which comprises agriculture, livestock breeding, forestry, extractive agriculture, fishing and aquiculture, the monthly value computed in cash is considered (market value), that the person usually earned. In the case of the remuneration in products or goods from the seasonal production, it is considered the monthly average value, real or estimated (market value) that the person usually earned.

For a person with a job but not at work and payed by a social security institution, the monthly gross income that he/she usually earned is regarded as the benefit in cash (illness benefit, work injury benefit, etc) relative to the reference month is considered:

For the employee, the monthly income usually earned excludes the amount that does not belong to a continuous nature (annual bonuses, premium pay for overtime, annual participation on profits, 13th salary, 14th salary, advanced payment of the 13th salary share, among others) and does not consider the occasional discounts (absence, advanced payment of the 13th salary share, eventual harm caused by enterprise and so on).

Monthly income effectively earned from employment

The income effectively earned from employment in the reference month is regarded as the one the person earned in fact that month.

For the remuneration in products or goods, from the activity which comprises agriculture, livestock breeding, forestry, extractive agriculture, fishing and aquiculture, the value in cash of the remuneration the person really used or withdrew in the reference month is considered.

For a person with a job but not at work and payed by a social security institution, it is considered the gross income effectively earned in cash (for instance, illness benefit and work injury benefit) during the reference month.

For the employee, the income effectively earned in the reference month includes all the extra gains (annual bonuses, delayed salary, premium pay for overtime, profit sharing, 13th salary, 14th salary, advanced payment of the 13th salary share) and considers all the occasional discounts (absence, advanced payment of the 13th salary share, eventual harm caused by enterprise, among others).

For the own-account worker and the employer, the income effectively earned in the reference month includes extra gains (annual bonuses, annual profit sharing, etc) and considers all the occasional losses (payment of eventual harm to the enterprise, for example).

J) Hours of work

Hours of work are those in which the person: works at the work place; works out of the work place doing tasks related to his/her occupation; stays in the work place available to perform his/her tasks; is waiting but cannot do his/her tasks because the machinery is out-of-order, due to accident, lack of material or assignment of tasks; it also includes the time dedicated to the preparation, maintenance and cleaning of work instruments; the necessary preparation to begin his/her tasks or elaborates controls, chronograms, reports and forms related to the job, including the ones due to legal obligations; or when he/she rests during short periods of time in the work place, including the moments to drink water, coffee or beverage. It is not included as hours of work, the time spent on the way from home to work and on meals. Since 2001, the number of hours usually worked and effectively worked have been investigated.

Hours usually worked per week

The hours usually worked per week are those the person used to dedicate to work. The hours usually worked per week refer to a typical period of work and should not be confused with the usual work hours, once the last ones refer to contractual conditions, which may not portray a typical situation of work.

For a person who had a job in which the hours of work did not vary according to the period of the year, it is considered the worked hours that demonstrated a week in which no exceptional situation altered the ordinary duration of work (sickness, vacation, holiday, time reducing, etc). For a person who had a job in which the usual duration of worked hours was different regarding the period of the year (such as seasonal activities), the hours usually worked that portrayed a typical week of the period of the reference week are considered. For a person who had a job in which the worked hours were very irregular, without a typical situation, the average of worked hours is considered. For a person who started to work in the reference week or in a recent period, it is considered the hours when he/she was waiting for work or intended to work are taken into consideration.Hours effectively worked during the reference week

The hours effectively worked are those the person in fact dedicated to work during the reference week.

Time – related underemployment in the reference week in all jobs: person who, during the reference week, effectively worked less than 40 hours in all jobs and was available to work more hours during the period of 30 days, counted from the first day of the reference week.

K) Classification of occupation and economic activity

Classification of occupations

Since 2001, PME uses the adaptation of Brazilian Classification of Occupations - CBO for household surveys called CBO-Domiciliar. For international comparisons, it is necessary to use a conversion table for the Uniform International Classification of Occupations (CIUO-88).

Classification of economic activities

Since 2001, PME uses the adaptation of Classificação Nacional de Atividades Econômicas - National Classification of Economic Activities – CNAE to household surveys called CNAE-Domiciliar. CNAE is compatible with the International Standard Classification of Industries (ISIC - 3) in two digits.

The individual information was coded in 5 digits, major disaggregation of CNAE-Domiciliar, however the micro data will only identify the first two digits regarding the analysing process of imputation used.

Once PME is a household sample survey, the indicators monthly published show information about activities grouped in eight categories. We should highlight that only seven out of them can guarantee some level of precision for the estimates

• Extrative and transformation industry and electricity, gas and water production and supply;
• Construction;
• Trade, repair of motor vehicles, personal and household goods, and retail sale of fuel;
• Services rendered to enterprises, rental and real estate activities, and financial intermediation;
• Education, health and social work, public administration and defense and compulsory social security;
• Activities of private households as employers Household Services;
• Other services - Hotels and restaurants; transport, storage and communications; refuse disposal; activities of membership organizations; recreational, cultural and sporting activities; personal services and
• Other activities - these are the activities that do not fit the categories above

L) Activity rate during the reference week - percentage of economically active population during the reference week in relation to the population aged 10 or over.

M) Unemployment rate during the reference week - percentage of unemployed population during the reference week in relation to the economically active population this week.

N) Employment rate during the reference week - percentage of employed population during the reference week in relation to the economically active population this week.

The Monthly Employment Survey covers the resident population in the metropolitan areas of Recife, Salvador, Belo Horizonte, Rio de Janeiro, São Paulo, and Porto Alegre since 1980. The geographic areas have been restricted to the urban areas mentioned above.

Since 2001, the published results refer to the population aged 10 and over. Previously, they used to refer to the population aged 15 and over.

PME is a continous survey with data collection throught the four weeks of each month.

PME uses a probabilistic sample of housing units, stratified and conglomerated in two levels, by metropolitan area of survey coverage.

In each city there is a selection of the primary sampling units (PSUs) and later of the secondary sampling units (SSUs). The primary sampling units of the survey are the census sectors, while the secondary sampling units are the housing units.

A systematic sample of households is drawn in the second stage.

The selection of sectors is made the systematic sample with proportional probability in relation to the total number of occupied private housing units obtained in the last Demographic Census. In the present methodology, a selected sector remains in the sample for all the decade or until the selection process is remade, unless the list of housing units of the sector finishes. In that case, the sector is substituded by another one with similar characteristics.

The listing operation for each selected PSU is done to enable the selection of housing units. It is characterized by the formulation of an updated register which enables the localization, identification and quantification of the housing units comprised by the selected area for the sample. The listing is annually updated or, exceptionally, whenever the selected area must be replaced.

After the selection of PSUs is made and based on the updated listing of housing units, the selection of household units is carried out by means of a simple systematic sample, considering 16 household units per area.

PME has aspects of a self-weighting sample design in each of the metropolitan areas. This implies that, due to the natural growth ordecline of the area, verified on each listing, the quantity of housing units to be selected may increase or decrease. However, since from 2001, a treatment for non-interview was included, the self-weighting characteristic is lost at the moment of the estimate of indicators.

The natural growth of the area does not interfere with the probability of its selection. However, significant increase of the number of housing units in certain areas may harm the quality of the estimators produced by the survey.

In order to isolate the significant growths, a “Register of Projects of New Constructions” is carried out in all municipalities of the metropolitan area, whether the areas were selected for the sample or not, identifying the area of New Constructions (N.C). The selection is carried out according to the sample fraction of each area.

Estimation Method

Considering the sample design of the PME, it is possible to say that the expected expansion factor is constant for all the housing units or persons of the sample in the metropolitan area, corresponding to the inverse of the sample fraction. However non-interview adjustment is made by PSU, in a way that the weights for all interviwed households are ajusted to account for ineligible selected occupied sample households for which no information was obtained because of absence, refusals or unavailability of the repondents for others reasons.The estimation procedure is based on ratio estimation. The weights with treatment for non-interview are adjusted again by a multiplying factor calculated independently for the metropolitan areas. This factor corresponds to the population projection for the reference period of the survey, divided by the sample estimation for the population using the weights adjusted for non interview.

Sample Rotation Scheme

The sample of housing units of the survey is distributed for each month, by the four reference weeks of the month. So, the result of the month is obtained by the average of these four reference weeks.

The monthly samples are not independents. The data collection follows a methodology in which each selected housing unit is interviewed during four consecutive months, it is not surveyed during the following eight months and, after this period, it is interviewed again during four months, and it is finally excluded from the sample.

It is important to point out that, during the period (12 months) in which the housing unit is in the sample, if the family moves out and another family moves into the housing unit, the information will be obtained from the new family for the period.

The PME monthly sample is subdivided into 8 rotation groups. Every month, 25% of the sample of housing units is replaced, following a scheme of rotation groups and panel. Each panel corresponds to a set of housing units and the rotation groups are the set of areas. Thus, for each month in a couple of consecutive years, 50% of common share of the sample is assured.

The PME rotation scheme represents the attempt to obtain gains in:

• Variance on the estimates of month-to-month changes, due to the fact that three-fouths of the sample are repeated in consecutive months.
• Variance on estimates of year-to-year changes, once half of the sample is the same for the same month in consecutive years.

Based on this sample rotation scheme, there are, in any month, housing units being surveyed from the first to the eighth time. In the present methodology, a selected area remains in the sample for a decade or until a new selection process is carried out provided that the list of housing units is not finished, in this case, the area is substituted by a new area that has similar characteristics.

The data obtained upon the PME are converted into desired indicators and estimates by means of the following steps: collection, coding, editing, imputation and analysis.

The surveys are distributed along the four consecutive weeks of each month, according to a pre-established collection timetable and decentralized polling. Every member of the household responds to an electronic questionaire in the selected household units, by personal interviews.

First of all, the interviewer obtains socio-demographic information from each member of the household and then their data on education and employment for every person aged 10 and over.

The ideal situation would be having each person providing data on himself/herself. However this may lead to the return of the interviwer several times, which would increase the time of this specific survey and increase the cost as well. In this case, the alternative is to collect data from one member of the household eligible to do so.

After each survey returns to the Database, the Supervisors check the integrity of the job and decide whether the ending of the survey will be proceded, which indicates the material is ready for the continuation of the process and for the decentralized polling. Then, the Supervisors carry out the assisted codification process on the described data that have not been automatically coded.

When the codification process ends so does the decentralized polling of the month. The data from each metropolitan area are sent to the Headquarters so that they are joined in a sole Database, which corresponds to the beginning of the centralized analysis process of data consistency.

The consistency process of the respondents´ answer registers begins with the input, with a software installed for the interview. This software is programmed to only admit the registers of valid values on the items with option of pre-defined answers. It includes the follow-up questions of the interview considering the correct flow of items, depending on the answer to previous questions.

After the transfer of the collected data to the database,as the information collection for each month of the survey ends, the set of information, by parts of the questionnaire, is submitted to a validation process with automatic correction of the inconsistent registers.

The part of general characteristics of residents goes through a deterministic imputation process, while a probabilistic imputation process through DIA (Detection and Automatic Imputation) system is applied on the parts of characteristics of education and labor. DIA is a software developed by the National Institute of Statistics - INE, from Spain, which enables the analysis and imputation of qualitative data. It is based on the Fellegi Holt methodology, with some changes to deal with systematic errors.

DIA is composed by two independent subsystems:

1. The probabilistic imputation subsystem based on rules of analysis (EDITS) which express situations (code combination of various variables) that are not acceptable in the same register.
2. The deterministic imputation subsystem based on deterministic imputation rules (RIDS). These rules express unacceptable or invalid combinations among the codes of several variables and specify the solution to be given to such situation.

The probabilistic imputation system works according to the following principles: to respect the original distribution of variables and to keep most of the original information.

The imputed register will be consistent, that is, it will respect all the specified rules.

The sampling data (originated from the corresponding estimators to the adopted sampling model) are expanded so that the desired estimates can be calculated.

At first, the analysis consists of the statistical analysis of the quality of the estimates and, then, of the establishment of the table plan for the press release.

Studies have been carried out in order to implement the imputation of income variables, by means of the regression tree technique and for the treatment of seasonal variations.