A SYNTHESIS OF KEY INSIGHTS AND LESSONS LEARNT

  • Slides: 10
Download presentation
A SYNTHESIS OF KEY INSIGHTS AND LESSONS LEARNT FROM COUNTRIES WHO HAVE USED REGISTERS

A SYNTHESIS OF KEY INSIGHTS AND LESSONS LEARNT FROM COUNTRIES WHO HAVE USED REGISTERS & ADMIN DATA IAN COPE

COUNTRIES USING REGISTERS IN CENSUS 2010 ROUND Registers only • Nordic Countries - Denmark,

COUNTRIES USING REGISTERS IN CENSUS 2010 ROUND Registers only • Nordic Countries - Denmark, Norway, Sweden, Finland • Austria, Belgium, • Slovenia, Turkey, Iceland Registers and Sample surveys • Spain, Germany, The Netherlands, Poland, Switzerland, Israel Registers and traditional census • Czech Republic, Latvia, Lithuania

COMMON FRAMEWORK FOR REGISTER AND COMBINED CENSUSES

COMMON FRAMEWORK FOR REGISTER AND COMBINED CENSUSES

REGISTER BASED CENSUSES UNIFIED IDS ACROSS SYSTEMS • A key enabler for the statistical

REGISTER BASED CENSUSES UNIFIED IDS ACROSS SYSTEMS • A key enabler for the statistical use of register data records is unified identification systems (IDs) across different register sources • Those countries with register based censuses all have Person ID systems and National Population Registers • with the same person ID used consistently across registers • In these countries unique identifiers are widely used for other registers like the business register and the registers on addresses, buildings and dwellings. • Without unique IDs used across registers it is more difficult to link different registers • Although countries such as the UK and New Zealand have linked successfully without a unified ID • But this is less certain and more complex

LINKAGE OF INDIVIDUALS BETWEEN REGISTERS • Models used include: • Linkage in the clear

LINKAGE OF INDIVIDUALS BETWEEN REGISTERS • Models used include: • Linkage in the clear (e. g. a database holding identifiers and characteristic variables), and • Trusted Third-Party (TTP) linkage models • The Trusted Third-Party linkage model is the most common: • Identifiers linked between registers and the pop register by TTP; a statistical identifier is created, attribute data extracted from source register • Attribute data linked using the statistical identifier • ONS/Stats NZ use this approach – linking on Name, Date of Birth, Address etc • TTP - records from individual registers linked together without personal identifier information being included in the data file • thereby preserving confidentiality and privacy

KEY REGISTERS Three key administrative base registers: • Register on persons (population register) •

KEY REGISTERS Three key administrative base registers: • Register on persons (population register) • Foundation for a pop register and ID system is a civil registration system • Critical that investments are made in these systems to ensure that they are of good quality, universal, continuous, permanent and compulsory. • Address Register (real estate, buildings and dwellings) • Business register With three associated statistical base registers: • Address is the link between the population/property registers; and business/buildings registers – giving location of residence and work • Other census variables come from registers on health, jobs, educational activities, tax, wages and salaries etc • but they are not base registers • In Nordic counties almost all specialised registers use the same ID key as the relevant base register - linking is thus normally straightforward

STEP-BY-STEP DEVELOPMENT The development process is the same across all countries using registers or

STEP-BY-STEP DEVELOPMENT The development process is the same across all countries using registers or admin data: 1 Ensure legislation in place to allow sharing of admin data with the NSO 2 Ensure that there is public trust/support for the use of registers/admin data 3 Develop positive relationships with register/admin data owners 4 Develop a linkage model 5 Produce an Address Register – align to concept of household in census 6 Produce a Statistical Population Register 7 Produce regular population estimates for local areas 8 Different subject matter statistics were developed, tested, and published • e. g. education, employment

STEP-BY-STEP DEVELOPMENT 9 Introduce register-based variables in the census when quality good enough 10

STEP-BY-STEP DEVELOPMENT 9 Introduce register-based variables in the census when quality good enough 10 - When statistics developed for all areas relevant for censuses, a totally register based census could be conducted 11 Developments in IT enable the time span between establishing admin registers and their use in official statistics to be reduced • Nordic countries took 30 years to develop fully register based censuses • Austria made transition from traditional census to a full register census in 10 years • with strong political, government and public support 12 Even today, the process will take time • Building systems and teams; quality assurance; setting up reporting procedures (may not work perfectly at first); etc

 • It’s important to start – you will gain experience and learn lessons

• It’s important to start – you will gain experience and learn lessons • Will take time to put in place data supply agreements, Mo. Us etc. • Take politicians and the public with you – sell benefits and protections • Business Registers likely in place; focus on Address Register and Population Register • Determine next priority – what high value admin data available?

A SYNTHESIS OF KEY INSIGHTS AND LESSONS LEARNT FROM COUNTRIES WHO HAVE USED REGISTERS

A SYNTHESIS OF KEY INSIGHTS AND LESSONS LEARNT FROM COUNTRIES WHO HAVE USED REGISTERS & ADMIN DATA IAN COPE