Benchmark and Base year revisions:Countries's experience

• Compilation of the benchmarks

• Procedure

• Issue on rebasing

• The changes

• Challenges

• Way forward

• Conclusion

A. is normally a culmination of years of significant data collection around the same accounting period including:

1. A household consumption survey
2. Employment survey
3. Informal/ the non observed economic survey
4. Business establishment census/ survey
5. Agriculture (crop and livestock) census/survey
6. NPISH survey
7. Population census
8. Administrative source data, external trade statistics

B. Updating of the activity/product classification to the desired ISIC revision, that is Rev. 4

C. Updating to the latest desired methodology (SNA) given the required data and resource availability:

1. Updating to the Work in Progress approach (SNA 93) for agriculture and construction activities. Some of the necessary SNA 93 principles had been applied during the previous rebase (Uganda's case)
2. Treatment of expenditures on R&D as GFCF rather than IC
3. Calculation of FISIM and its allocation (Cameroon, Madagascar and Uganda)

1. Measurement of the biological resources under GFCF (Uganda)
2. Estimation of the production of non life insurance by the adjusted claims (Cameroon, Madagascar and Uganda)
3. Treatment of goods exported for processing and the processing fee (Madagascar, planned for the next rebase)

Step 1: Classification of all survey/census estimates to CPC and ISIC

Step 2: Compiling SUT for each products (product balance )

Step 3: Compilation of institutional sector SUT or informal vs formal

Step 4: Feed the classified data directly into the SUT

Step 5: Diagnostic of aggregates for which the source data are not reliable, insufficient

Step 6: Preliminary balancing, manual or automated

Step 7: Second, third, etc. balancing until the final SUT.

A. Choosing the rebase year:

• Year when economy is stable (politically and economically)
• Reference year for most surveys/census (Benchmark year)

B. Number of years for which GDP can be recompiled:

• Depends on data requirements suitable for the updated methods
• Resource constraints (financial and HR)

• Rebasing ensures better coverage of the economy than in the past and normally results in an increase in the current price GDP level

• Here is the GDP at current market prices for Uganda:

• From 1990 to 2005

Lessons learnt:

• Not to use borrowed software to balance the SUT, unless it was developed for your country (Uganda).

• Unanticipated delays due to manual balancing and high staff turnover

• Experience sharing between experienced national accountants and the new comers.

A. Madagascar:

• A balanced SUT of 2007 using SNA 93 and ISIC Rev.4 (in current price of 2007) has been done. GDP increase is about 60%
• 2007 will be the new base year (data availability, before crisis)
• Preparing to rebase to 2007 this year (2015) or in 2016 but also to conduct national business establishment survey

B. Cameroon:

• Estimation of non market output for the regional central Bank and military systems under GFCF
• Improve the coverage of household consumption

C. Uganda:

• Preparing to conduct 2015 reference year estimates for the next rebase.

Important to:

1. Include estimation for activities that are known to evolve prior to the next rebasing.
2. be frequently inform the major data users and the general public at different stages about the critical on-going issues during the benchmarking including the “expectations”.
3. include 3-5 years of revised series in the rebased series to affirm changes in the base year. Series should include some years before the base year.
4. Routinely make data revisions and methodological updates when It is possible
5. Have the same base year as the year for the benchmarks