The results of the analysis have presented some interesting findings that highlight the value of the data and the investigation. The investigation has shown that specifically targeting the PBLCI households when collecting expenditure information and using that information to derive expenditure weights within the expenditure classes gives different index results in 24 of the 85 expenditure classes compared with the existing method. During the five year period (March 2006 to March 2011) covered by the analysis, the impact on the published quarterly and annual percentage changes would have been minor positive and minor negative effects at different times. The effect on the index level would have been a marginal upward trend.
The investigation also shows that there are some goods and services that are a significant part of the PBLCI households' expenditure that are not currently represented in the published PBLCI and that there are retail outlets at which these households shop that are not currently represented in the published PBLCI.
This investigation has shown that the current published PBLCI could be improved upon by making use of expenditure information relevant to PBLCI households at the more detailed levels. Judging from the results of the analysis, the impact on the published PBLCI figures would be minimal. If suitable detailed expenditure information can be regularly sourced, the ABS will make use of this information to improve the index. Sourcing the appropriate data will require additional funding.
There are potential benefits from introducing 'new' goods and services into the PBLCI and collecting prices from a wider range of retail outlets but this has not been analysed. The ABS is considering this as part of its wider programme of improvement in its price index concepts, data sources and methods.