Use of administrative data in statistics Nordic experiences Kaija Ruotsalainen UN World Data Forum 16 18 January, Cape Town, South Africa
Summary: registers as an input for statistics - Use of the registers for statistics has a long history in the Nordic Countries 1980 Population and Housing census totally based on registers in Denmark first in the word Finland followed 1990, Norway and Sweden 2011 - However, registers were used as a part of statistical processes widely before that, e.g in Finland population and taxation registers were used already in1970 census Population structure were compiled based on registers in 1970 s in all Nordic countries - Registers can be used in statistical processes in different ways: statistics based on a one register (population structure, vital statistics) combining several registers together (population and housing census) Combining registers with survey data (LFS, EU-SILC etc) Sampling frames (LFS, EU-SILC etc) Quality control 2
Summary: registers as an input for statistics The year of establishing registers/introducing registers in census statistics in Nordic countries Register data Denmark Finland Norway Sweden Established First Estab- First Established First Established First used in lished Census used in Census used in Census used in Census Population 1968 1981 1969 1970 1964 1970 1967 1975 Business 1975 1981 1975 1980 1965 1980 1963 1975 Dwellings 1977 1981 1980 1985 2001 2011 2008 2011 Housing conditions 1977 1981 1980 1985 2001 2011 2008 2011 Education 1971 1981 1970 1975 1970 1980 1985 1990 Employment 1979 1981 1987 1990 1978 2001 1985 1985 Family 1968 1981 1978 1980 1964 1980 1960 1975 Household 1968 1981 1970 1975 2001 2011 2011 2011 Income 1970 1981 1969 1970 1967 1980 1968 1975 Totally register-based Census 1981 1990 2011 2011 Source: Register-based statistics in the Nordic countries, UNECE 2007 3
Summary: requirements - Quality frameworks for statistics like UN Fundamental Principles of Statistics, European Statistics Code of Practice etc - Most of them underline the NSIs right to access the administrative sources to produce the official statistics - This is one of the basic precondition of the utilisation of admin data for statistics - What else is needed? Method to combine data from different sources: best situation is if there is a unique id widely used in all administrative registers Understanding the benefits Co-operation inside and outside NSIs 4
Summary: output - Important to keep in mind: Administrative data is collected for administrative purposes not in principal for statistics BUT: it is useful to use also for statistics if the quality is good enough Administrative data is used in the administrative decision making processes at the individual level BUT: when administrative data is used for statistics, it is processed to be as statistical data, it is confidential and used only as an aggregated data in planning and decicion making It is not possible to use in the administrative decision making processes at the individual level 5
New possibilities for research and statistics production (1) - Statistics usually provide cross-sectional information on a variable at a given point in time, such as population number or the number of people in gainful employment. - On this basis we can see to what extent these figures have changed. - The register system offers the added advantage of allowing us to identify the individuals behind these changes: who has got a job, who has completed a degree. - Changes can be monitored by linking unit data from consecutive years. 6
New possibilities for research and statistics production (2) - Traditionally, the most important regional unit in statistics has been the administrative area. - However, administration is dynamic and keeps changing => may occur difficulties to keep up these changes - The building-based code system with its coordinates has provided a solid foundation for reliable and flexible statistical areas. - Despite major changes in administrative areas, it is still possible to produce time series for different regions. - The adoption of map coordinates for buildings has also make it possible to define more flexible statistical areas. 7
New possibilities for research and statistics production (3) For instance: - Calculation of accessibility (workplaces, services) Distance to work, school, voting place - Flow statistics employment flows student flows - Longitudinal researches Data from 1970, 1975, 1980 and 1985 Censuces and annually data from year 1987 in the Census Data Warehouse Data over 7 million persons (Population of Finland 5,4 million in 2015) 8
Some examples
Flows between different activity groups: Employed 2012-2013 10
Employment rate, % Employment rate by birth cohorts 100 90 80 70 60 50 40 30 20 10 0 20 25 30 35 40 45 50 55 60 64 Age 1950 1960 1970 1980 11
Kaija Ruotsalainen kaija.ruotsalainen@stat.fi