ICP Regional Coordinators Meeting Washington - DC, April 20-21, 2011 [06.02]
ICP Regional Coordinators Meeting
Washington - DC, April 20-21, 2011
[06.02]
1. Objectives
2. HHC Validation Overview
4. Global validation
Outline
Submission and Validation
3. HHC Price Data
5. Schedule of Data and Metadata Submission
1. Agree on validation process and procedures
in general
2. Agree on transmission of data and metadata
3. Agree on validation process and stages during
Global validation
Objectives
• Intra-country
• For regional items (incl. CGL)
National validation
NCAs
• Inter-country
• For regional items (incl. CGL)
Regional validation
NCAs, RCAs
• Inter-country across regions
• For CGL items and regional PPPs
Global validation
NCAs, RCAs, GO
Validation stages
HHC Validation Overview
National Level Regional Level Global Level
• Quarterly Prices and Metadata• Annual Prices and Metadata
Original or Edited• Quarterly Prices and Metadata• Annual Prices and Metadata• Validation Tables• PPPs by the ICP Classification
Original or Edited
• Quarterly Prices and Metadata• Annual Prices and Metadata• Validation Tables• PPPs by the ICP Classification
Validation Queries on
• Quarterly Prices and Metadata• Annual Prices and Metadata
Validation Queries on
HHC Validation Overview
1 2
Quarterly Price Data and Metadata
HHC Price data
• Validated national quarterly average prices (whole, location types, outlet types)
• Validated sub-national quarterly average prices• Importance indicators
• Validation tables (Number of price quotations , Min prices, Max prices, Min/Max price ratios, Variation coefficients, Outlier indicators)
• Local currencies, quarterly average exchange rates of local currencies
• Metadata• RCs to GO: PPPs by the ICP Classification
Annual Price Data and Metadata• Validated national annual average prices (whole, location types, outlet types)
• Validated sub-national annual average prices• Importance indicators
• Validation tables (Number of price quotations , Min prices, Max prices, Min/Max price ratios, Variation coefficients, Outlier indicators)
• Local currencies, annual average exchange rates of local currencies
• Metadata• RCs to GO: PPPs by the ICP Classification
National Coordinators (NCAs)Intra-country and sub-national
• Quarterly data
Tasks
• Initial data validation
• Annual data
• Computation of national annual average prices
• Validation of national annual average prices (sub-regional and temporal)
Regional Coordinators (RCAs)Inter-country
• Quarterly data
• Annual data
Tasks
• Validation of average prices (QT/DT) at BH and aggregated levels
• Helping in the validation of observations, if needed• Helping in the computation and validation of national annual averages, if needed
• Validation of preliminary PPPs
• Temporal validation of average prices and preliminary PPPs
• Statistical data validation (observations and average prices)
NCAs and RCAs
• Sub-national validation
• Temporal validation (across quarters)
Global Office (GO)Inter-country across regions
Global Office
• Helping in the validation of regional prices, if needed
• Validation of linked regional PPPs
• Validation of CGL prices
• Inter-country validated Quarterly and Annual data
• Inter-country validated Quarterly and Annual PPPs
Tasks
• Helping in the validation of preliminary PPPs, if needed
• Providing validation documentation
• Validation of reference PPPs
• Validation of regional validation tables
Global Validation
• Validate data globally (GCL prices, PPPs, importance and reference PPPs)
TAG agreements• Use weighted CPD as an official calculation method
• Use CAR approach for linking regional PPPs
• Compare results of different regional linking options
Heston/Dikhanov (CAR)Linking factors are
weighted harmonic means of ratios between global and regional PPPs
Weighted CPD Model recognizes that representative or important products usually have a larger
expenditure share than do non representative or less important products. Approximate weights can be applied; for example, a
representative product could receive twice the weight of a non representative product. A weighted least squares method is used to
estimate the PPPs.
Regional average prices and PPPs should remain open for review until the global validation of
the core prices is completed
Global Validation
• This has to be taken to account by both NCAs and RCAs
Global validation of data will result in validation queries and proposals for edits
• Relating, regional data should not be published before
• Timing may be an issue
Validation of GCL Prices
Global Validation
• Validation of PLS (for each level of aggregation)
• (Regional validation within inter-country validation)
• Validation of GCL prices deflated by regional PPPs
• Between region BH PPPs
Validation of regional PPPs
Input: inter-country validated quarterly and annual average prices and quarterly and annual regional PPPs
• Validation of importance (straight and deflated prices)
• Validation of Direct vs. Indirect PPPs (for each level of aggregation)
• Global validation - matrix of 180+ countries (item, BH, aggregate)
• Validation of reference PPPs
Global Validation
• BH PPPs – implausible cases
Validation rules• Prices – variation or high/low CUP-ratios (residuals)
• Importance – if low/high CUP-ratio item + percentages
• Global office flags outliers for RCAs and NCAs
Final decisions on validation actions
• Aggregated PPPs – implausible cases
• Validation group will make final decision on used prices
• Data may be used in regional comparison without being used in global comparison
Use of data in regional and global comparisons
• Reference PPPs – implausible cases
13
Quarterly Price Data and Metadata & Validation Table
Annual Price Data and Metadata & Validation Table
Q1
2011 2012 2013
Q4
Q3
Q4
Q3
Q4
Q3
Q2
Q1
Q2
Q1
Q2
Q1
PPPs by ICP Classification & Metadata
(Preliminary and Final)
Q2
Q3
Q4
Q1
Q2
Q3
Q4
P
A
A
NCs to RCs
RCs to GO
F
Schedule of Data Submission
14
THANK YOU