1. The hidden side of successful story – implication of wide use of administrative data sources at national statistical institutes Metka Zaletel, Irena Križman Statistical Office of the Republic of Slovenia

2. IAOS 2008, Shanghai Introduction Reduction of burden in statistics can be achieved by –minimising demands of users (number of variables, frequency of data collection, excluding some of the reporting units) and –by using already collected data, either exchanging them within the statistical system or using administrative data with modern technology and organisation.  New challenges, new tasks and new demands at the national statistical institutes

3. IAOS 2008, Shanghai Wide usage of administrative sources Necessary preconditions: –Good legal framework –High level of cooperation with other important players: Other authorised producers Owners of administrative data sources and registers Institutions collecting data for multiple purposes

4. IAOS 2008, Shanghai Implication of use of administrative sources to data processing Legal – administrative tasks –Agreements –Basic centralized check of data Centralized control and editing of administrative data source –an expert with the deep knowledge of the content of the source –Follow up of changes in the content and the structure Application of the administrative source in a survey –adjustment and adaptation of the methodology of the source –Prioritization of different data sources –Cleaning when combined

5. IAOS 2008, Shanghai Implication of use of administrative sources to data processing

6. IAOS 2008, Shanghai New challenges for human resource management and training The problem of employees who have been performing routine work for last 20 years, and are not needed anymore New training programmes should be developed for the survey managers and their assistants Number of employees dedicated to a certain statistical survey might increase at some surveys because of complexity of new tasks

7. IAOS 2008, Shanghai Cost and burden – case study Survey on Travels of Domestic Population is a case of simple data collection: sample size about 6000 households per quarter, sample design – stratified systematic sample, interviewing mode is CATI, quarterly survey. EU-SILC is a case of complex data collection: sample size about 12000 households, sample design rotating panel, stratified two stage sample, 40% of variables collected by CAPI or CATI questionnaire, 60% of variables collected by different administrative sources, annual survey.

8. IAOS 2008, Shanghai Cost and burden – case study (2)

9. IAOS 2008, Shanghai Cost and burden – case study (3) The two surveys are not completely comparable because of –different sample sizes, –partly different interviewing modes and –slight difference in number of observed variables. Data processing should be comparable if administrative sources didn’t play an important role in EU-SILC processing.

10. IAOS 2008, Shanghai Cost and burden – case study (4) More than 7500 hours were spent on EU-SILC in 2007 Almost 3000 hours were spent on travelling survey. The difference is caused –partly by larger number of variables, –mostly by extensive use of administrative variables and a lot of problems while combining 15 different administrative sources by the field questionnaire. We estimate is that about 1500 - 2000 hours is spent just for the data matching, cleaning and prioritizing, what means around 20.000 EUR extra costs within the office. What would happen on the other side (during the fieldwork) if the administrative data were not used at all: burden of households would increase for 100% or for almost additional 40.000 EUR per year.

11. IAOS 2008, Shanghai Conclusions Implications of wide usage of administrative sources within NSIs are serious and usually neglected by governments. Decrease of response burden by using administrative sources generates new demands and needs within NSIs. Final recommendation: to use as much administrative data sources as possible, but handle them with caution within the NSI with the emphasis on the human resource management and training.