The usability of administrative data for registerbased censuses

The usability of administrative data for registerbased censuses Eric Schulte Nordholt

Outline (1) – – – – Data collection Data sources History of the Dutch census Publication of the Dutch Census 2011 Variables of the Census 2011 Characteristics of the Dutch census Conditions facilitating use of administrative sources 2

Outline (2) – – – Two hundred years of censuses Censuses in the UNECE region Current and future work Educational Attaiment File Imputing the Educational Attainment File Discussion 3

Data collection – Centralised in one data collection division (registers and surveys) – Efficiency and professionality – Using the same infrastructure for social and economic statistics as much as possible – Collection strategy (preferred order): – register data – sample surveys – Only existing sources have been used for the Census 2011 – No Census 2011 questionnaires! 4

Data sources For the Census 2011: – Registers – – – – Population Register (automated population registration system) Jobs file (all employees) Self‐employed file (all self‐employed) Fiscal administration Social security administration Pensions and life insurance benefits Housing Register – Survey – Labour Force Survey 5

History of the Dutch census (1) Traditional censuses till 1971: Ministry of Home Affairs Statistics Netherlands 1829 1849 1869 1889 1909 1930 1960 1839 1859 1879 1899 1920 1947 1971 • Unwillingness (non-response), privacy concerns, reduction expenses • No more traditional censuses 6

History of the Dutch census (2) Virtual censuses from 1981 onwards: Register-based, survey information added, no census questionnaires 1981 1991 • Population Register and surveys • No official publication, no internal consistency, no regional results 2001 2011 • System of social statistical datasets (SSD) • European Census Act of 2008 • 2011 Census published in European Census Hub 7

Publication of the Dutch Census 2011 Census book: http: //www. cbs. nl/en‐GB/menu/themas/dossiers/historische‐ reeksen/publicaties/archief/2014‐dutch‐census‐ 2011‐ pub. htm? Languageswitch=on Census tables (of 32 European countries): https: //ec. europa. eu/Census. Hub 2/ 8

Variables of the Census 2011 (1) Persons: 1. Place of usual residence 2. Location place of work 3. Locality size 4. Sex 5. Age 6. Legal marital status 7. Current activity status 8. Occupation 9. Industry 10. Status in employment 11. Educational attainment 12. Country/place of birth 13. Country of citizenship 14. Year of arrival in the country 15. Place of usual residence one year before 16. Housing arrangements 4, 1% Employed 8, 9% Unemployed Under 15 years 16, 6% 49, 1% Pension or capital income recipients Students (not economically active) 17, 5% 3, 8% Homemakers and others 9

Variables of the Census 2011 (2) Household variables: 1. Household status 2. Family status 3. Type of family nucleus 4. Size of family nucleus 5. Type op private household 6. Size of private household 7. Tenure status of households Housing variables: 1. Type of living quarters 2. Occupancy status of conventional dwellings 3. Type of ownership 4. Number of occupants 5. Useful floor space and/or Number of rooms 6. Density standard 7. Water supply system 8. Toilet facilities 9. Bathing facilities 10. Type of heating 11. Dwellings by period of construction 12. Dwellings by type of building 10

Characteristics of the Dutch census – – – – – Relatively cheap Short production time More socially acceptable than the Census 1971 All tables are numerically consistent Some missing information in the survey part Dependent on registers (availablility, timeliness) Comparisons over time possible International comparisons possible Census data continuously available 11

Conditions facilitating use of administrative sources – – Legal base (Statistics Act) Public approval (‘Big Brother is watching you’) Cooperation among authorities (mainly government organisations) Comprehensive and reliable register system (administrative versus statistical quality) – Unified identification system (preferably unique ID‐numbers) 12

Two hundred years of censuses Inhabitants and household size Number of inhabitants (x mln) / Mean houshold size 18 16 14 12 10 8 6 4 2 0 1829 1839 1849 1859 1879 1889 1899 1909 1920 1930 1947 1960 1971 1981 1991 2001 2011 Census year Number of inhabitants Mean household size 13

Censuses in the UNECE region 14

Current and future work Preparing the Census 2021: – New European implementing regulations (based on UNECE recommendations) – Preparatory projects at Statistics Netherlands: ‐ Methodology (estimating occupation tables) ‐ Software ‐ Level of education attained (imputing the Educational Attainment File) ‐ Housing variables ‐ Population grid squares (legal base, protection of census data) 15

Educational Attainment File – Complex integration process of microdata from LFS and examination registers – New version containing also information on private education institutions available since 2016 – About 60 % of the records have information on highest level of education attained – Weighting to known marginals of the population for statistics on level of education – Better quality and more detailed education tables than before when only LFS information was used 16

Imputing the Educational Attainment File In this project: – Find a good imputation model for the Educational Attainment File (logistic regression model) – Produce a set of hypercubes of the Census 2011 again, now by using the imputed Educational Attainment File – Develop a set of quality indicators (basis for decision how detailed the future census publication will be) – Plan for highest level of education attained in the Dutch Census 2021 (also of interest for other countries) 17

Thank you very much for your attention! Holland in the winter 18

Discussion Are there questions or remarks? 19