Lars Thygesen

1
International Trade Expert meetingWhats going
on at OECD statistical information management

• Lars Thygesen

2
Data collection from Member countries
3
Modes of data collection

• standard questionnaire
• send what youve got
• web queries from web site

4
Quality

• transmit on the day of publication
• available on web site for download
• follow specified format
• correct translation of codes

5
Highlights from a survey

• numerous cases of missing data (national
  accounts)
• short and long delays
• some errors and re-sendings

6
What can be done?

• Short term Remove obstacles?
• Long term New methods?

7
What is SDMX?

• Statistical Data and Metadata eXchange
• develop and employ more efficient business
  processes for exchange and sharing of statistical
  information and metadata
• Standards

8
And why is OECD interested?

• Questionnaires
• heavy reporting burden
• similar questionnaires for other IOs
• timeliness
• error prone
• heavy burden on OECD

9
What we hope to gain

• more efficient reporting, i.e. less burden on
  countries as well as on OECD
• automated procedures which are less error prone
• more timely reporting and, consequently,
  dissemination
• enhanced coordination with other international
  organisations

10
Aligned Projects

• Exchange of national accounts data between the
  OECD and Member countries (NAWWE)
• Joint UN/OECD Foreign Trade Statistics project
  (Comtrade)

11
NAWWE
Data Flow
OECD Web Server
Web Application
SDMX Based XML Schemas
XML Files Validated By Schema
XML Data Files
Navigation
Structural XML
lt?xml version"1.0"?gt ltOECD.STAT
xmlnsxsd...gt ltValuegtCANADAlt/Valuegt lt/Datagt

Upload to SNA Database or Output to Other Format
Defines Dimensions and Data Structure
lt?xml version"1.0"?gt ltOECD.STAT
xmlnsxsd...gt ltValuegtAUSTRALIAlt/Valuegt
lt/Datagt
Copied XML Files
XML Files Copied to Server
XML Data Files
Content
12
OECDs new Statistical Information System
13
Overall goals

• User friendliness
• One-stop database
• Easy access
• Common look and feel of all products
• Efficiency
• Common tools
• Easy access across OECD
• Quality
• Coherence of concepts data
• Timeliness
• Good metadata
• Review of processes
• Flexibility
• Separate storing, process, presentation

14
OECD.Stat

• The central repository ("warehouse") where
  validated statistics and related metadata
  destined for publication and/or sharing are
  stored

15
SIS Architecture
Production
Storage
Dissemination
Work flow
16
Metadata principles

• All statistical data being shared or
  disseminated, internally or externally, must have
  appropriate metadata
• Statistical metadata must be consistent
• All metadata must be created only once

17
Quality problems Recommen- dations
SIS Migration Plan
Quality Review Statistics X
18
The End