Improving Coordination within the National Statistical System Dennis

Improving Co-ordination within the National Statistical System Dennis Trewin

Centralised v Decentralised Systems • All statistial systems have some degree of centralisation or decentralisation • Statistical co-ordination is important in both • Principle 8 of the Fundamental Principles of Official Statistics “Co-ordination among statistical agencies within countries is essential to achieve consistency and efficiency in the statistical system”

Why is co-ordination important and becoming more important? • Users are seeking greater consistency across data sources • Users want to access all data (irrespective of the source) that is available using the internet • Governments increasingly wanted to work in a connected way • Statistics should be based on sound methods • Efficiency may be improved by sharing statistical infrastructure

Australian Legislation to support Co -ordination • Provides Australian Bureau of Statistics with co-ordination power to improve compatibility, reduce respondent burden and improve statistical standards • Statistical Clearing House has been created • Relies on co-operation not cohesion • Advisory Council responsibility covers the whole national statistical system

New Zealand Legislation to support Co-ordination • Minister must approve all statistical collections • Has delegated that function to Government Statistician except for the most important statistics • NSO leadership role reinforced by recent review • In practice, a lot of reliance on protocols, etc except for the most important statistics • Advisory Committee established to oversight the national statistical system • Statistics NZ given archival function

Global Trends • Many countries have made or are making changes to legislation to improve the coordination of their national statistical systems (eg Japan, Mexico, United Kingdom)

Statistical Leadership • More statistical expertise exists in NSOs – best positioned to provide leadership on statistical issues • Leadership is more than co-ordination • Requires relationship management skills, entrepreneurship, and a capacity to add value to statistical work outside the NSO

Some practical ways of improving statistical leadership • Developing standard classifications and support resources for use by all (eg geography) • Statistical frameworks, standards and definitions for use by all • Manuals of good practice and supporting training programs • Agreed protocols for the operation of the NSS • Directories of statistical sources • Development of infrastructure for general use • Establishing centres of statistical expertise for particular subject matters • Information Development Plans

Information Development Plans Key features - Jointly agreed by all stakeholders especially the data providers - Focus is on a particular field of statistics - Identifies existing data (from across the NSS) that are relevant to the field - Identifies highest user priorities for improvements - Identifies plans for future action including responsibilities, timetables, etc

OECD Quality Reviews • Reviews applied to all Divisions not just the Statistics Division • Had strong support of Secretary-General • Sought views of users • Made constructive suggestions for improvement • Seen as adding value by the producers of the statistics

Barriers to Improved Co-ordination • Sufficient priority being given to statistical issues • Resistance to change • Statistical skills and facilities • NSO resourcing • Unwillingness to put data in the public domain

Discussion Points? • How can I reduce resistance to NSOs playing an active role in the NSS? • How can the NSO add most value?