Download presentation
Presentation is loading. Please wait.
Published bySara Payne Modified over 9 years ago
2
© Statistisches Bundesamt, VI A Statistisches Bundesamt The new method of the next german Population census Johann Szenzenstein, Federal Statistical Office, Germany
3
© Statistisches Bundesamt, VI A Statistisches Bundesamt Folie 2 The most important element of the German census method is the use of a combination of administrative registers and surveys as data sources The main data sources are: Population registers Employees registers Housing census (postal survey) Sample survey
4
© Statistisches Bundesamt, VI A Statistisches Bundesamt Folie 3 The main features of the Population Registers Population registers (PRs) are kept decentrally by the municipalities (about 13 500 municipalities). Every person living in a municipality is legally obliged to register in the PR There is no central population register. There is no ID-number stored in the PRs. Most of the characteristics and their values stored in the PRs are standardised by federal law.
5
© Statistisches Bundesamt, VI A Statistisches Bundesamt Folie 4 Additional variables stored in the PRs which can be used to group persons together to form families/households family name, first names and date of birth of the spouse family name, first names and date of birth of the children only for children (aged under 27 years): family name, first names and date of birth of the parents address of the previous place of residence date of moving to the current dwelling
6
© Statistisches Bundesamt, VI A Statistisches Bundesamt Folie 5 Civil servants, army personnel and judges Coverage of the employees registers A. Employees Registers maintained by the Federal Employment Agency (FEA) All employed persons subject to obligatory insurance contributions All persons registered as unemployed at the labour administration All persons attending a vocational training programme of the German Federal Employment Agency B.Other administrative employees registers Not covered: self-employed persons and contributing family workers (= 10% of economically active population)
7
© Statistisches Bundesamt, VI A Statistisches Bundesamt Folie 6 A. Census variables which can be obtained from the employees registers Economic variables current activity status (employed, unemployed) occupation industry time usually worked (full/part time) status in employment (apprentice, wage earner, salaried employee) place of work Educational variable educational attainment Variables used for exact record linkage to the CPR records family name, surnames date of birth place of residence
8
© Statistisches Bundesamt, VI A Statistisches Bundesamt Folie 7 Housing variables for buildings: period of construction, type of building, number of dwellings for dwellings: occupancy status, tenure status, number of rooms, living floor space, kitchen, bathing/toilet facilities, type of heating, main type of heating, monthly (gross and net) rent. Variables which can be used to generate private households (these variables have to be provided for each occupied dwelling): the number of occupants the names of one or two occupants and their dates of moving to the dwelling. Census variables collected from the owners of the buildings by a postal survey.
9
© Statistisches Bundesamt, VI A Statistisches Bundesamt Folie 8 (Simplified) Model of a Register-based Census in Germany Housing census (questionnaires) (Technical) Standardisation + Merging Central Population Register (CPR) Checking for completeness + plausibility checks Records for buildings and dwellings Records for persons: economic variables Adjusted CPR; records for persons Checking for multiple entries + adjusting for overcounts Generating (private) households + adjusting for errors in the overcounts Linking CPR records to records from ERs Completed records for -persons -households -buildings Records for - persons - private households - buildings - dwellings Municipal Population Registers (Technical) Standardisation + Merging Employee registers
10
© Statistisches Bundesamt, VI A Statistisches Bundesamt Folie 9 Test surveys to prepare a register-based census in Germany (carried out for the reference day, 5 th December 2001) evaluating the quality of the population registers (overcounts/undercounts) testing the efficiency of different matching techniques to identify double (or multiple) entries in the (central) population register establishing the efficiency of the newly developed algorithm for generating households testing the feasibility of microdata linkage between different data sources without uniform ID-numbers evaluating the quality of housing census data collected from the owners of the buildings by a postal survey
11
© Statistisches Bundesamt, VI A Statistisches Bundesamt Folie 10 Test survey for evaluating the quality of the register data (register test) Sampling method :about 38 000 addresses (buildings) in 550 municipalities with about 250 000 dwellings and about 550 000 residents Procedure: requesting PR records for all persons registered at the selected addresses from the municipalities interviewing all households living in the selected buildings by enumerators comparing PR records with data from household survey Main objectives: estimating the number of overcounts and undercounts of the PRs for all 16 federal states and 4 municipality size classes
12
© Statistisches Bundesamt, VI A Statistisches Bundesamt Folie 11 Results of the register test Table 3.1 Overcounts and undercounts of the registers for persons at their main place of residence by municipality size classes. Municipality size class (inhabitants) Persons registered *) Including undercountsovercounts 1 000 percent1 000percent under 10 000 22 947.5303.61.3634.62.8 10 000 – 49 999 26 112.7348.41.3900.03.5 50 000 – 799 999 23 944.5509.32.11 175.74.9 800 000 and over 6 980.5207.13.0527.27.6 Total 79 984.91 368.41.73 237.54.1 ________________ *) Persons living in institutional households are excluded.
13
© Statistisches Bundesamt, VI A Statistisches Bundesamt Folie 12 Test survey for identifying the double entries in the PRs (double-entry check) Sampling method: all persons born on 1 st January, 15 th May, 15 th September or with an incomplete date of birth (about 21,2% of the population) were sampled Procedure: requesting PR records for all sampled persons from all municipalities (about 970 000 records) checking for multiple entries with 6 different matching techniques for identifying duplicates setting up unclear cases by phone, mail or by field interviews Main objectives: estimating the of number of persons with double entries in the population registers improving the matching procedures
14
© Statistisches Bundesamt, VI A Statistisches Bundesamt Folie 13 Results of the double-entry check Table 3.2 Overcounts in the CPR after checking for (permanent) double entries by municipality size classes (persons at their main place of residence). Municipality size class (inhabitants) Persons registered*) Including net Overcount**) cleared up by double-entry check remaining overcounts 1 000% % % under 10 000 23 071.0459.52.0149.90.7309.61.4 10 000 - 49 999 26 928.1643.42.5153.30.6490.11.9 50 000 - 799 999 24 839.1801.63.4139.30.6662.32.8 800 000 and over 7 342.0416.36.043.00.6373.35.4 Total 82 180.02 320.82.9485.50.61 835.32.3 _____________ *) Persons living in institutional households are included. **) Excluding “ temporary“ overcounts.
15
© Statistisches Bundesamt, VI A Statistisches Bundesamt Folie 14 Main conclusions from the test surveys: As census results (number of inhabitants) are used for the distribution of funds among the municipalities census results of high accuracy are needed. The test surveys have shown that inhabitant numbers deduced from adjusted PR cannot meet the required accuracy standards: the overcount- undercount rates vary considerably between small towns and big cities. It is therefore proposed to complete the tested model of a register-based census with a sample survey. This supplementary sample survey has to serve 2 purposes: Estimation of the number of overcounts and undercounts for each municipality (with an accuracy of at least +-1%, given a probability of 95%) Collection of additional census variables missing in the registers It is suggested that the sample survey should provide reliable results only for municipalities with 10 000 and more inhabitants (necessary sample size: about 6 mill. persons)
16
© Statistisches Bundesamt, VI A Statistisches Bundesamt Folie 15 Strengths and weaknesses of the new census method Strengths: the new data collection method will be much cheaper (estimated costs: EUR 336 million, of which costs for the housing census: EUR 250 million) than a traditional census (estimated costs: EUR 1 020 million) the new approach will involve a much smaller response burden on the citizens (in total about 27 million respondents) than a complete enumeration of the population (about 82 million respondents) Weaknesses: the main disadvantage of the new approach is that it cannot guarantee full census information for small municipalities and reliable census results for local units below the municipality level.
17
© Statistisches Bundesamt, VI A Statistisches Bundesamt Folie 16 Thank you very much for your attention
18
© Statistisches Bundesamt, VI A Statistisches Bundesamt Folie 17 Census variables stored in the Population Registers Demographic variables: sex age legal marital status country of citizenship country/place of birth religious affiliation Geographic variables place of usual residence
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.