Statistical quality in the European Central Bank: experience to date

1
Statistical quality in the European Central Bank
experience to date and way forward
Catherine Ahsbahs, Violetta Damia Gérard Salou,
Caroline Willeke
Q2006 Cardiff, 24 - 26 April 2006
This presentation reflects the views of the
authors and not necessarily those of the ECB
2
Overview

• General considerations
• Quality assurance procedures in the various steps
  of the statistical production chain
• Quality monitoring and reporting
• Quality related procedures
• Adherence to principles of good governance and
  quality
• Conclusion and outlook

2
3
! General considerations
3
4
! General considerations (II)
4
5
Quality assurance procedures in the
statistical production chain
5
6

• Development of new statistics

6
7
Development of new statistics Merits Costs

7
8

• Data Collection

• Secondary data collection (i.e. NCBs ? ECB)
• Data exchange between the ECB and partner
  institutions using SDMX/EDI (GESMES/TS)
• Standard data model (data set structures
  based on the SDMX data model) standard format
• Common nomenclatures
• Quality controls on data reception
• Syntax, semantics (e.g. code values), verify
  expected (or not) series
• Exchange of metadata also in SDMX-EDI

8
9

• Compilation data quality assessment

• Quality checks
• reception status
• integrity of the series no missing data
• revision checks
• comparison against previous data
• advance release calendar
• revision studies (e.g. life-cycle)

9
10

• Compilation data quality assessment (II)

• Quality checks (cont)
• consistency accounting checks
• Smonths year
• Sflows stock of previous stock of current
• re/classifications revaluations
• with other data sources (BSI-b.o.p.,
  b.o.p.-external trade, etc.)
• plausibility checks
• univariate analysis

10
11

• Compilation data quality assessment (III)

• Quality checks (cont)
• production of data quality report
• analysis of data at an aggregate level
• Automation
• minimum manual interventions
• Harmonisation across data series
• use of shared concepts code lists to the max.
  possible extent

11
12

• Confidentiality protection

• Protection of confidential statistical
  information
• Legal regime based on
• Council Regulation (EC) No 2533/98 Guideline
  ECB/1998/NP28
• ECB statistical systems prevent any further
  dissemination of information reported as
  confidential statistical information (CSI)
  care is also taken to prevent indirect deduction
  of CSI.
• Prior to amendment of Reg. No 2533/98, interim
  measures for the exchange of confidential
  statistical information within the ESCB for
  statistical purposes
• Practical implementation

12
13

• Data accessibility Dissemination policy

• Equal and easy access via various dissemination
  means (Monthly Bulletin, Statistics Pocketbook,
  Website)
• Eurosystem Joint Dissemination
• Statistical Data Warehouse
• General public - September 06
• Exchange of information with intern.
  organisations
• Assistance to the users

13
14
Quality monitoring reporting
14
15
Quality related procedures
Risk process management
15
16
? Principles of good governance and quality

• ECB statistics directorate general mission
  statement
• The Directorate General Statistics strives to
  provide the ESCB and the general public
  statistical information fit for use in the
  decision-making of the ECB
• CCSA principles
• ECB commitment to adhere to Principles governing
  international statistical activities
• Way forward
• Developing a quality framework for the ECB
  statistical function

16
17
? Principles of good governance and quality
(II)

• Taking into account
• Specific ECB institutional environment
• Application of the Eurosystem Mission Statement
  to the ECB statistical function
• IMF DQAF Code of Practice of the ESS OECD
  Quality Framework

17
18
Conclusion and outlook

• Next steps
• Improve the procedures
• documentation analysis to identify good
  practices among the various production areas
• Investigate the usefulness of expanding
  quantitative quality indicators to other
  statistical areas than b.o.p./i.i.p.
• development of an ECB quality framework
• Further enhance interactions with users
• Share good practices with other international
  national organisations

18
19
THANK YOU VERY MUCH FOR YOUR ATTENTION!