EXECUTIVE SUMMARY

1. In March 2016 the ABS announced it was conducting a review of the Retail Trade Price Index (RTPI) to assess the concepts, data sources and methods used to compile the RTPIs, as well as the performance of the RTPIs.

2. The review has been completed following extensive review of the methodology, analysis of results using different data sets, consultation with users and reviewing international experience.

3. The review has put forward a number of key recommendations:

• The ABS will publish a quarterly national RTPI for use in the stage of production (SOP) framework, and more generally by external users, to assist in understanding the transmission of price changes throughout the economy
• The ABS will produce supply use product classification (SUPC) national retail trade price indexes, as annualised measures, to meet National Accounts requirements
• State and outlet data will no longer be collected from providers which is a major reduction in the number of data items to be collected

4. Feedback from users is being sought on these recommendations and other general comments about the RTPI.

5. Subject to feedback from stakeholders, the ABS will progressively implement the recommendations from this review from the December quarter 2016. All of the recommendations will be fully implemented by the September quarter 2017.

BACKGROUND

6. The ABS has published an experimental RTPI series presenting national data commencing December quarter 2003. State level data and data for selected product groups have been published since March quarter 2008.

7. The current design of the RTPI is as described in 6427.0.55.007 - Experimental Producer Price Index for the Output of the Retail Trade Industry, Australia, 2013

THE CURRENT SET OF OUTPUTS

8. The set of outputs published in 6427.0 - Producer Price Indexes. Australia are:

• National Retail Trade price index, percentage change from previous quarter, and percentage change from same quarter last year
• National Retail Trade price index for selected product groups, percentage change from previous quarter, and percentage change from same quarter last year
• Retail Trade price index by State/Territory, percentage change from previous quarter, and percentage change from same quarter last year

9. Due to timing constraints in the collection of data, RTPI outputs are published around 25 days later than other Producer Price Indexes and the RTPIs are not currently included in SOP.

THE CURRENT DATA REQUIREMENTS FOR PRODUCING RTPIs

10. Providers are required to provide quarterly Sales and cost of goods sold (COGS) data for selected products (SUPCs).

11. One important concept of a price index is that the quality of the good or service being priced is held to a constant quality. The quality of the distribution service provided by a retailer is difficult to measure. Any attempt to measure the quality of the distribution service needs to consider characteristics such as outlet opening hours, numbers of checkouts, floor space, general ambience, temperament of staff, ease of parking, range of products on offer, proximity to other stores etc. While these characteristics may not lend themselves to ready measurement it has been assumed they would be linked to the specific outlet providing the products. Accordingly data providers have been asked to provide data for the three largest outlets in each state and territory the business operates.

12. Providers have also reported on their total State/Territory sales for the selected products which is used to weight indexes to produce State aggregates.

THE PROPOSED METHOD AND DATA REQUIREMENTS

13. The review found controlling for quality by monitoring outlet characteristics is difficult. Furthermore analysis of data collected since 2008 has shown that the use of outlet data, rather than using the businesses aggregate data, has an insignificant impact on the RTPI even at the most detailed supply use product classification (SUPC) level. This conclusion was supported by other international experience. The ABS has therefore decided to stop collecting outlet data. In future, data providers will only report for their total business sales and cost of goods sold (COGS) by SUPC.

14. The review has also determined there are two main uses of the RTPI

• Use by the ABS's National Accounts Benchmarks section for the compilation of annual supply use tables.
• For use as a national retail price index in the stage of production (SOP) framework, and more generally by external users, to assist in understanding the transmission of price changes throughout the economy.
  

15. These uses do not require a State/Territory dimension to the indexes. Furthermore, analysis of current State RTPIs has shown that for major States, with robust samples, there is a strong coherence between the indexes and a single National index should be sufficient for all practical purposes.

16, To date although SUPC indexes have not been published, they have been aggregated and published as eight product groups. These product groups will no longer be produced or published.

17. The ABS will continue to produce the RTPI for major SUPCs and in aggregate for a national RTPI. Only the national RTPI will be published quarterly. SUPC indexes will be produced quarterly and made available as annualised indexes to National Accounts. State and outlet data will no longer be collected from providers which is a major reduction in the number of data items to be collected.

18. RTPIs can exhibit considerable quarterly variability, particularly at SUPC level, and this reflects international experience. There are many real world and methodological factors which can contribute to this variability. National Accounts will use an annualised index for their purposes.

PROCESSES

19. The review has identified a number of improvements that can be made to existing processes. Many of these are the result of the major reduction in the number of data items collected and needing to be processed.

• the current web forms will be simplified to reflect the reduced data requirements and this should make it much easier for providers to enter data and reduce the risk of data transcription errors
• samples will be reviewed to ensure indexes are robust at the SUPC level. This may result in either reductions or increases in sample given that there is now no stratification of sample by state although good national coverage will still be required.
• editing and processing systems will be revised to improve significance editing capabilities. Where possible automatic detection and correction of data anomalies will be introduced.

20. The aim of these improvements is to enable a fit for purpose national RTPI to be published and incorporated in the SOP framework concurrently with the rest of PPI. Completion of SUPC indexes can be done post publication and this may result in some revisions to the national RTPI which will be implemented in the following quarter.

21. The review also looked at a number of longer term options involving different collection strategies and processing systems. These will be considered in the context of the ongoing five year program to transform the ABS infrastructure, systems and processes used to produce official statistics.

USER FEEDBACK

22. The ABS welcomes feedback on these recommendations. Readers wishing to comment or request further information should contact Tom Lay, Director, Prices Branch on (02) 6252 6157 or via email on prices.statistics@abs.gov.au.

23. Please provide feedback by COB Tuesday 20th of December 2016.