1. The Australian Bureau of Statistics (ABS) Consumer Price Index (CPI) is a robust measure of household inflation. The CPI is compiled quarterly and the expenditure weights, primarily sourced from the ABS Household Expenditure Survey (HES), are updated every 6 years. The methods, data sources and compilation frequency of the CPI have served Australia well over many decades.
2. While the CPI is a well-respected ABS output, recent developments in methods and the availability of electronic transactions (scanner) data presents new opportunities to enhance the CPI.
3. The ABS Prices Branch has commenced a research program to enhance the Australian CPI. This project will re-examine the traditional approach to collecting data and compiling the CPI, and in doing so, potentially creates new avenues to enhance the Australian CPI.
4. Consultation with CPI users, particularly during the 16th Series CPI review, has highlighted areas of the CPI that could be enhanced. These areas can be grouped into: (1) Frequency of Expenditure Class weight updates; (2) The future role of transactions datasets; (3) Compilation frequency; and (4) Other enhancements to various components of the CPI.
5. The remainder of this paper describes the implementation of CPI enhancements undertaken to date; and outlines future research work to examine methods and approaches to enhance the Australian CPI.
Part 1: Work undertaken to date
Transactions (scanner) data
6. The ABS significantly increased the use of transactions data in the CPI March quarter 2014 publication. The approached used to compile the CPI using transactions data is described in the September quarter 2013 Consumer Price Index feature article(footnote 1). In short, the ABS implemented a 'price replacement' approach using product unit values. The use of product unit values is internationally endorsed in the International Monetary Fund (IMF) CPI Manual(footnote 2). The approach takes revenue and quantity data by product over the period of interest, i.e. a quarter, and calculates an average price. This average price is then combined with weights to calculate price indexes for Expenditure Classes and the total CPI. The major benefit of this approach compared to the traditional point-in-time pricing is that a more representative price for each product is used to compile the CPI. This approach enhances the accuracy of the CPI.
7. Transactions data are currently used to price products that represent approximately 25% of the weight of the CPI. A list of Expenditure Classes that contain products that are priced using at least some transactions data can be found in Attachment 1.
8. While the unit value approach to pricing sampled products in the CPI is an enhancement, more can be done to maximise the use of transactions data to compile the CPI. Part 2: Research program to enhance the CPI examines how a greater use of transactions data can improve the accuracy and potentially the frequency of the CPI.
Part 2: Research program to enhance the CPI
9. Research work has commenced to examine methods and approaches to enhance the Australian CPI. The ABS will progressively release the results of this research for stakeholder consultation. The research will include input and advice from international price index experts.
Frequency of weight updates
10. The CPI Expenditure Class weights are updated every 6 years, primarily using data from the ABS Household Expenditure Survey (HES). Below this level, products and product weights are updated on a rolling basis each quarter.
11. Research is advanced to utilise Supply Use Product Code (SUPC) Household Final Consumption Expenditure (HFCE) data as the main data source to more frequently update Expenditure Class weights. The aim is to annually update these weights. The ABS will, in the short term, publish detailed information outlining the proposed approach, methods, concepts, benefits and a potential path to implementation which will be used for consultation with CPI users.
12. The ABS is investigating methods to maximise the use of the transactions datasets. That is, to move beyond the current 'replacement' strategy.
13. The research program will consider expanding the CPI product samples and using quantity data to weight the lower levels of the index; as well as assessing internationally developed methods to compile price indexes from transactions datasets.
14. The ABS has commenced this research into the greater use of transactions datasets. The ABS will publish detailed information of the methods and strategies being assessed for consultation with users.
Producing a monthly CPI
15. The ABS CPI is published quarterly. This is based on a sample of prices across a variety of frequency schedules (e.g. weekly, monthly, quarterly, annually). The current CPI collection schedule, by Expenditure Class, used to compile the CPI can be found in Attachment 2.
16. The ABS has previously provided an estimate of costs to produce a monthly CPI of equivalent quality to the current quarterly index . This estimate was based on the traditional approach to collecting data and compiling the CPI. The availability of transactions and administrative datasets creates new opportunities to produce the CPI at lower cost, and potentially at higher frequency. Action is underway to obtain a greater number of transactions datasets from retail businesses to further assess the feasibility of compiling a more frequent CPI. As more transactions datasets become available, the production of a monthly CPI becomes more feasible.
17. In May 2015 the Government announced a major investment in the ABS to modernise and transform the ABS' ICT business systems, operations and processes. This investment provides the infrastructure required by the ABS to continue delivering timely and quality economic, environmental, demographic and social statistics for Australia. This work will modernise and streamline systems and processes supporting the production of the CPI. This work will also reduce the future cost of producing a monthly CPI.
18. The ABS will need to prioritise the production of a more frequent CPI against other statistical programs, noting the initial and ongoing cost to the ABS. The ABS recognises that a monthly CPI is the prevailing international standard, but this does need to be considered in the context of the broader ABS work program and available resources.
19. The ABS will publish detailed information relating to the compilation/publication frequency of the CPI in due course.
Other enhancements to the CPI
20. The ABS Information paper Outcomes of the 16th series Australian Consumer Price Index review, December 2010, (cat. no. 6469.0) identified a selection of enhancements and opportunities that could be pursued by the ABS.
21. This research program will work progressively through these outcomes and publish information for user consultation. One example is the potential role of transactions datasets to expand price collection beyond capital cities.
22. Location information obtained through transactions and administrative datasets is being assessed for this purpose. In addition, the ABS will investigate the possibility of using these datasets to compile spatial indexes. This research is in its formative stages. Further information will be published in due course.
23. The ABS has, and will continue to consult with peer National Statistical Offices (NSOs) and international price index experts as part of this research program. Peer review of the research outcomes will occur.
Part 3: Where to from here
24. The ABS will progress the research program and release detailed information for user consultation over the coming year.
1 Feature Article: The use of transactions data to compile the Australian Consumer Price Index, Published in Consumer Price Index, September quarter 2013, (cat. no. 6401.0)
2 IMF Consumer Price Index: Theory and Practice