Statistical Data Collection IMF Statistics Department

1
Statistical Data Collection IMF Statistics
Department

• OECD Experts Group Meeting
• April 2004

2
Topics

• Scope
• Objectives
• Techniques Tools
• Challenges
• Successes
• Looking Ahead SDMX

3
Scope why?

• Data for surveillance
• IMF Articles of Agreement require the provision
  of data by members
• Article IV Obligations regarding exchange rate
  arrangements
• Article VIII Obligations of members section
  5(a) specifies twelve categories of national data
  as the minimum necessary for the effective
  discharge of the Funds duties
• Data for publication
• IFS, IFSY, BOPSY, DOTS/Y, GFSY

4
Scope who?

• 180 member countries, 4 or more agencies per
  country (500 agencies)
• National Statistical Offices
• Central Banks
• Ministries of Finance
• Customs offices, etc
• Regional agencies
• BCEAO, BEAC, ECB, ECCB, Eurostat etc
• International agencies
• BIS, ILO, OECD, UNSD etc

5
Scope what?

• Real Sector statistics (NSOs, CBs)
• National accounts, prices, production, labor,
  trade
• Financial (CBs)
• Money and banking, interest rates, share prices
• Fiscal statistics (MOFs)
• Government financial operations
• External Statistics (CBs, Customs)
• Balance of payments, IIP, trade, international
  reserves, exchange rates, CPIS

6
Objectives

• Automate data processing
• Accuracy, timeliness, efficiency
• Scalable and manageable approach
• Provide security for data exchange (encryption
  and authentication)
• Reduce reporting burden on members
• eliminate duplication
• use standard formats

7
Techniques

• Countries
• System of country correspondents
• Report Forms (paper or electronic) including
  current data from our database
• Mix of formats
• Data sharing with other organizations
• Mix of formats data feeds, CDs, diskettes

8
Tools

• Integrated Correspondence System (ICS)
• Secure, web-based electronic data reporting
• 160 countries 800 individuals 375 forms
  monthly
• Countries query our databases (Freq, Per)
• Automated identification of new and revised data
• BOPSTA and GESMES/TS (limited)
• Other electronic media (limited)
• Country web sites (very limited)
• Web services (limited)
• Possible pilot with OECD, e.g., IFS, Industrial
  Production

9
Challenges

• Country-specific data sets hinder harmonization
  and automation
• Availability of technology varies widely among
  data providers
• Improve collection of structural metadata (time
  series and observation-level)
• Minimize the number of solutions deployed
• With 500 different provider agencies , we
  cannot sustain a country/agency approach

10
Successes

• Recent migration from mainframe and re-engineered
  business processes
• GESMES/TS feeds from ECB, BIS, others
• Key family facility for automating database to
  database reporting, i.e. GESMES/TS
• Moving towards standardized report forms in
  financial statistics
• Collecting data from centers instead of member
  countries, e.g., Industrial Production (OECD),
  labor (ILO), etc..

11
Looking Ahead SDMX

• SDMX information model SDMX-ML Key families.
• Countries post data on web in SDMX format, STA
  grabs
• the dataset(s) it needs.

• What wed like to see in place
• Security (encryption and authentication)
• Bi-directional exchange (query ability?)
• Exchange just new and revised data
• Notification when new data are posted
• Questions
• Maintain role of centre organization key
  families?
• Expand exchange of metadata