Back to top
National reference metadata

Romania

Reference metadata describe statistical concepts and methodologies used for the collection and generation of data. They provide information on data quality and, since they are strongly content-oriented, assist users in interpreting the data. Reference metadata, unlike structural metadata, can be decoupled from the data.

For more information, please consult our metadata website section.

Close

Census 2011 round (cens_11r)

National Reference Metadata in Euro SDMX Metadata Structure (ESMS)

Compiling agency: Institute of National Statistics

Need help? Contact the Eurostat user support


Short metadata
Full metadata

Data given in this domain refer to the 2011 Population and Dwelling Census carried-out in Romania in October 2011. Data were collected by enumerators hired by local public administrations and processed by National Institute of Statistics.

2011 Population and Dwelling Census applied the requests of European Regulations and United Nations and Eurostat Recommendations for the 2010 Censuses of Population and Housing. In the same time, the Romanian dedicated legislation was applied.

The essential aim of 2011 Population and Dwelling Census was to obtain and offer basic and good quality statistical information for governmental policies in the economic and social fields in the benefit of human development.

More specific, the goals of the census was to obtain more detailed and good quality statistics about the number and territorial distribution of population, about it’s demographic and socio-economic structure and information about population households,  building stock and living conditions of population.

Based on this very important information, it will be possible to analyze the dynamics of several demographic and social phenomena, both at national and international level.

31 March 2014

The 2011 Population and Dwelling Census had as observation units: persons, dwellings, buildings for human habitation, other buildings in which at least one person has usual residence, households and families.

Statistical population is referring to the persons who had the usual residence in Romania for a period of 12 months or more in the census reference date or who established the usual residence inside Romania for less than 12 months but had the intention to stay at least 12 months.

The Census was carried-out in all Romanian territory, without any exception.

20 October 2011

Not aplicable

After all sets of questionnaires were delivered to the County’s Technical Secretariat of Population and Dwelling Census, the data entry activity begun.

Before data entry starting, special category of census staff - the persons in charged with coding – performed the activity of codification (manual) for all variables to open questions using predefined classification and nomenclatures: citizenship, ethnic group, mother tongue, religion, localities (usual residence, place of birth, domicile, previous residence, place of work), country of previous residence, school of highest level graduated, school attained, occupation and industry.

Data entry was done using operators.

For the linkage between different questionnaires in a specific category the identification code was used. For example, for linkage between individual questionnaires of household’s members and household and dwelling questionnaires, the components of this identification code had the following meaning:

-         Folder number (census sector code)

-         Order number of the building within the folder

-         Order number of the dwelling within the building

-         Number of the LC form within the folder

-         Number of the SC form under which the dwelling was recorded

-         Number of G form of the dwelling

-         Person’s order number

For identification of missing individual records in census microdata database and administrative sources used in the indirect data collection the personal numerical code (CNP) and persons’ addresses were used.

For each kind of questionnaire, a special list of logical conditions were prepared and included in the data entry and data checking software. There are around 1000 logical conditions which checked:

-         presence or absence of obligatory variables’ answers;

-         logical correlation between several variables;

-         outliers and special cases;

-         normal limits for numerical variables;

-         logical correlation between variables in different questionnaires (for example between information regarding de facto or de jure marital status and the existence of the husband/ partner in the same household) and so on.

A set of checking was done in the same time with data editing. The second set of checking was done after data editing, in the phase of data validation. These activities were done at each county’s level.

When data editing and data validation were finished, a second package of operations were done at INS level, namely:

-         invalid personal numeric codes (CNP);

-         looking and cleaning for double records (two records for the same person) in the same file (P, TP or PPI file);

-         missing records;

Based on lists with errors, these errors were solved at county’s level too.

After processing the individual forms, the under-registration found in 2011 Census’s provisional results processing stage was confirmed. Therefore, methods have been applied to ensure the census data completeness, using both the method of indirect collection from administrative sources and statistical techniques for imputation of data. There were 1,183 thousand persons not enumerated in the Population and Dwelling Census of October 2011 (hereinafter, PHC 2011) and identified at the administrative data sources level. The administrative data sources found to have useful information for finalizing the results of 2011 Census are contained in:

  • The National Register of Personal Data - RNEP – managed by the Directorate for Persons Record and Databases Management;
  • Statement on obligations to pay social security contributions and income tax, and the nominal records on insured persons - D112 – managed by the National Agency for Fiscal Administration;
  • Record of Employees - IM – managed by the Labor Inspectorate;
  • CNPP database –  managed by the National House of Public Pensions (CNPP);
  • CNAS database –  National Health Insurance House of Romania (CNAS);
  • Tax Registration Statement / Statement for individuals who carry out economic activities independently or liberal professions –  D070 - managed by the National Agency for Fiscal Administration
  • Record of beneficiaries of state child allowance, family allowance and help the guaranteed minimum aid – managed by the National Agency for Payments and Social Inspection (ANPSI)
  • Database of students enrolled in the 2011-2012 school year – managed by the Ministry of Education.

The starting point of the indirect collection from administrative sources procedure was the comparison of records in the 2011 Census database (data recorded on the individual forms completed by the enumerators during the field collection stage) with the existing entries from the Population Register database. The set of records found in the Population Register and missing from the PHC 2011 database (recorded on all P, PPI or TP forms[1]) were searched in other administrative sources for the month of October 2011 and the preceding and subsequent months of the same year, in accordance with the concept of residence in Romania for at least 12 months; this concept was applied during the 2011 PHC to measure the usual resident population. Just for persons identified in the administrative sources used, for which there is clear evidence that they were in Romania during the Census and most of 2011, a record was added (imputed) in the PHC 2011 database for which identifiers and values ​​ were subsequently filled for the Census variables. Thus, all recordings obtained from indirect collection have the same structure with the rest of records obtained from interviewing people during the field data collection period and refer to the same reference point, enabling the aggregation of information for the country’s entire usual resident population, whether or not interviewed by enumerators.

The minors pertaining to the identified adults (mother and father) for whom information were taken from the administrative sources above, as well as for adults (mother and father) who were counted during the PHC 2011, were searched and identified in the National Register of Personal Data.

Based on the information related to the address of the enumerated persons, data on dwellings were also completed and the households within those housing were reconstituted.

 The principle of item imputation was to use the best source for imputation and best criteria to find a good proxy for the real information. Therefore, item imputation can be classified in two main categories:

  1. Some variables were imputed from the administrative data sources where we found it; we consider this an indirect collection of data, not an imputation, because these data are information declared by individuals, so they represent valid values, not artificial ones (as it is the case of item imputation).
  2. The second category refers to statistical imputations. Depending on type of variables, we applied the following kind of item imputation:
  • for variables referring to individuals, we used the hot deck donor method;
  • for the qualitative variables related to dwellings, we used the method of most frequent value from the cell a specific dwelling is part of ;
  • for the qualitative variables related to dwellings, we used the method of cell’s average for that cell a specific dwelling is part of ;

The administrative data sources were used in cascade, one after one. The information found in administrative data sources were use for item imputation of census’s variables on total imputed records.

If the information was available for missing persons, the following variables were collected from administrative data sources by indirect data collection for imputed records: CNP (sex and date of birth derived from CNP), legal marital status, domicile, previous residence, citizenship, highest level of school graduated, educational institution the person is attending, current activity status, work time, occupation, employment status, industry and type of sector in which the person is working.

The automatic corrections applied solved the data inconsistencies, like for example:

-         age - highest level of school graduated – current activity status;

-         occupation – industry - type of sector in which the person is working;

-         age - educational institution the person is attending - highest level of school graduated;

-         age – sex - legal marital status ;

-         age – number of life-born children;

Due to special character of some variables and due to non-existence of this information in the administrative data sources, we leave the answer “Not available” for the following variables:

- year of marriage; year of entering the consensual union; number of live-born children (only for women); ethnic group; mother tongue; religion; difficulties in performing the current activity (in P questionnaire)

- dwelling type of ownership; number of rooms used for professional purposes; dwelling’s endowment with air conditioning installation; year of building’s construction; construction material for external walls of the building (in LC questionnaire)

We delete the records considered to be duplicates of the same person/ household or dwelling.

Duplicate CNP were classified in:

  1. pair records with the same CNP enumerated inside a specific county and
  2. pair records with the same CNP enumerated in two different counties
  1. the County’s staff had to identify the paper questionnaires for the two pair records and compare the content on paper with the information included in database.

-         If, from name and first name, date of birth and other information comparison they could conclude the two questionnaires referred to the same person, one questionnaire was deleted from the database;

-         Contrary, it was clear one questionnaire contained a valid and correct CNP and the second one contained a wrong one. In this case, the staff found the correct CNP for the second questionnaire, corroborating the information existing on the fulfilled second questionnaire;

-         If the correction of the CNP was not possible, it was made zero and the record remained in the database.

 

b. pair questionnaires with the same CNP enumerated in two different counties were firstly checked from the answers’ completeness point of view on the majority of variables. Elimination of one double record was done based on following criteria:

-         From 2 individual records 18 years old and over with current activity status as student, the record enumerated in an university town was kept;

-         From 2 individual records having fulfilled the locality in which the workplace is situated, one of them having workplace in the enumeration locality and the other in different locality, the record in which the work place was located in enumeration locality was kept;

-         From 2 individual records, one being allocated to an LC- Dwelling, building questionnaire and the other one being allocated to a SC questionnaire (collective living unit), the record from SC questionnaire was kept;

-         From 2 individual records, one being fulfilled for a present person in the enumeration locality and the other one being fulfilled for a temporarily absent person, the record for the present person was kept.

Once all members of a household were deleted, the record corresponding to these deleted records was deleted too.

 

Starting with the imputed individual records, households were determined based on address (all persons living in the same dwelling) and family relationships.

In some cases, information about family composition was found in Social protection administrative data (child allocation) and in Population register.

Two different cases were taken into consideration:

    1. for the imputed individual records associated with a dwelling already recorded in the census micro-data database – in this case the imputed record was included in the household already recorded and the family relationships were made over;
    2. for the imputed individual records associated with a dwelling (address) missing in the census micro-data database

The remaining imputed records, where no family association was found, were considered households with unrelated members and un-known type of household.


[1]  P – Person (present or temporarily absent); PPI – Person left for a long period of time (inside or outside the country); TP – person temporarily present .

Following the initial dissemination policy, there were three phase of data dissemination:

-       on 2-th of February 2012 as provisional results

-       on 24-th of August as preliminary results

-       on 4-th of July as final results.

27 months between data collection period and data disemmination through hypercubes.

From methodological point of view, all indicators are consistent and comparable across all geographical units: macro-regions, development regions, counties, localities.

The 2011 Population and Dwelling Census followed very closely the definition requested by Regulation no. 763/2008 and the United Nations and Eurostat Recommendations for the 2010 censuses of Population and Housing.

Non-derived topics

Deviation which could impair the EU-wide comparability of the data

- Place of usual residence

No deviation

- Sex

No deviation

- Age

No deviation

- Legal marital status

No deviation

- Country/place of birth

No deviation

- Country of citizenship

No deviation

- Previous place of usual residence and date of arrival in the current place; or place of usual residence one year prior to the census

No deviation

- Relationships between household members

No deviation

- Location of place of work

No deviation

- Current activity status

No deviation

- Occupation

No deviation

- Industry (branch of economic activity)

No deviation

- Status in employment

The category “Others” includes, besides “Contributing family workers” and “Members of producers’ cooperatives” a third category “Other situation”. It has no big impact on the category “Others”.

- Educational attainment

No deviation

- Ever resided abroad and year of arrival in the country (from 1980)

No deviation

- Relationships between household members

No deviation

- Tenure status of households

No deviation

 

 

Derived topics

 

- Total population

No deviation

- Locality

No deviation

- Household status

No deviation

- Family status

No deviation

- Type of family nucleus

No deviation

- Size of family nucleus

No deviation

- Type of private household

No deviation

- Size of private household

No deviation

Non-derived topics

 

- Type of living quarters

No deviation

- Location of living quarters

No deviation

- Occupancy status of conventional dwellings

No deviation

- Number of occupants

No deviation

- Useful floor space and/or number of rooms of housing units

No deviation

- Housing arrangements

No deviation

- Type of ownership

No deviation

- Water supply system

No deviation

- Toilet facilities

No deviation

- Bathing facilities

No deviation

- Type of heating

No deviation

- Dwellings by type of building

No deviation

- Dwellings by period of construction

No deviation

 Derived topics

 

- Density standard

No derivation; the density standard was calculated based on useful floor space in square meters to the number of occupants.

 

The 2011 Population and Dwelling Census measured only primary homelessness (roofless). Due to the complex system of questions needed to measure secondary homelessness, these questions were not included in the census questionnaires.

The homeless persons were enumerated in the places were the enumerators found them: on the street, on collective living quarters (hospitals, special City hall’s shelters, other living units) and a P-questionnaire was fulfilled.

Full comparability over time for geographical units at national, counties’ level and the majority of localities.

For a number of 110 administrative units (mainly from rural area) the comparability with the 2002 population and dwelling census data is not possible due to modification of the locality’s structure during the last intercensus period (some localities divided, some other merged).