SUMMARY AND CONCLUSIONS
4.1 The Australian CPI is a robust indicator of household inflation that has served Australia well for many decades. While the data sources and methods are well understood by users, there are particular aspects of the CPI that can be enhanced. In light of this, the ABS has embarked on a research program aimed at enhancing the Australian CPI. This paper focuses on one of the areas identified for enhancement, being the frequency of the EC weight updates.
4.2 The ABS currently updates expenditure weights at the EC level every six years using data from the HES. The availability of HFCE data from the National Accounts provides the ABS with an opportunity to update CPI EC weights more frequently. The ability to reweight the Australian CPI more frequently would have significant benefits to the user community. These include more representative expenditure weights that more accurately reflect consumer spending patterns, addressing stakeholder concerns following the 16th Series CPI review, coherence across macroeconomic statistics and, improved alignment with international standards.
4.3 The use of expenditure aggregates from the National Accounts is a recommended approach from the ILO in situations where the time interval between household surveys is large. Within Australia, HFCE data is available on an annual basis as part of the Australian System of National Accounts (cat. no. 5204.0). Practical challenges exist (due to scope and coverage differences) when using HFCE data for CPI weights. These have been examined in detail (in chapter 2) with treatments recommended.
4.4 The ABS has conducted empirical investigations with results demonstrating broad stability in weights at the group level when comparing HFCE (t-1, t-2 and t-3) and HES data, indicating both sources tell a coherent story of consumer expenditure patterns.
4.5 The ABS has produced three experimental HFCE price series over the period September 2005 to September 2015. The results reveal that all three experimental HFCE series reported lower average annual household inflation measures relative to the CPI. They were quantified as 2.54% (t-1), 2.58% (t-2) and 2.61% (t-3), compared to 2.67% for the CPI over the analysis period.
4.6 The empirical results support the theory that higher frequency re-weighting at the EC level captures consumers' substitution effects. When compared to a retrospective superlative index between the period September 2005 to September 2011, all three experimental HFCE series reported a lower estimate of substitution bias. This was quantified as 0.09% (t-1) per annum, 0.15% (t-2) per annum and 0.16% (t-3) per annum, compared to 0.24% per annum for the CPI. These results indicate that annually chain-linked HFCE price indexes capture a greater amount of consumer substitution when compared to the current re-weighting process (i.e. six yearly updates).
4.7 With respect to the available HFCE vintages, t-1 vintage is recommended as the preferred option. The t-1 vintage is preferred based on three main criteria; (1) it has the shortest implementation lag; (2) it is closest in proximity to a superlative index; and (3) it aligns with international practice.
4.8 While National Accounts annual HFCE data is historically revised, the Australian CPI is only revised in exceptional circumstances. These results show that the impacts of historical revisions to National Accounts HFCE data are small. This analysis shows that historically revised HFCE price indexes are very close to the t-1 vintage.
4.9 Since the HES is a major benchmark of the HFCE and CPI, a more frequent HES is still desirable.
4.10 The ABS will update CPI weights with HES data 2015-16 in December quarter 2017. The HES 2015-16 will also be incorporated into National Accounts HFCE data, ensuring consistency across macro-economic statistics.
4.11 After December quarter 2017, the ABS will move to update the CPI EC weights on an annual basis using the methods and data described in this paper.
4.12 The ABS will review its strategy of implementing HFCE weights in the CPI. Readers are invited to provide feedback on this paper. Specific points on which feedback would be appreciated are:
- proposed treatment of HFCE data to align with CPI;
- proposed treatment of geographic coverage differences;
- proposed alternate data sources to HFCE data; and
- maintaining the current quality of the SLCIs (using six-yearly weight updates).
4.13 Please provide feedback to email@example.com
outlines how the ABS will handle any personal information that you provide to us.
4.14 A timetable for feedback and consultation is as follows:
- 11 July 2016: Release and widely distribute this information paper summarising investigations into more frequently updating CPI EC weights
- July - September 2016: Consult with users regarding the contents of the information paper
- October - November 2016: Communicate outcomes of feedback and consultation process.