Your browser is not supported

Our website does not support the browser you are using. For a better browsing experience update to a compatible browser like the latest browsers from Chrome, Firefox and Safari.

Our submission to the 2013 CPI Advisory Committee

Our submission to the 2013 Consumer Price Index (CPI) Advisory Committee outlines the purpose and frequency of the CPI, incorporating retail transaction data as well as regional and subnational data and dissemination of CPI information.

Release date: 24/04/2013

Purpose of the CPI

We recognise there are many uses for the consumer price index (CPI) and that uses other than inflation targeting have different requirements of the index. A single CPI has the benefit of providing a reputable, objective and reliable measure of price changes facing households. As such, we seek to maintain the publication of the CPI as a single index. The current reliability and credibility of the CPI assists us to maintain stable inflation expectations, an important outcome sought by monetary policy deliberations.

We strongly support the continued use of the acquisition-based approach as the conceptual framework for the CPI. This approach matches the intuition that the price measured should be the one that people experience when purchasing a good or service. While there is merit in the use method, it is less intuitive and likely better suited to analytical series.

We believe the current index is sufficient to meet the wider needs of a majority of users. However, we also believe that Statistics New Zealand could publish a larger number of analytical series to meet demand for different constructs. Users currently have a wide range of series published alongside the official CPI that provide significant additional data and information. The provision of these analytical data has, in our view, enhanced rather than detracted from the overall usefulness of the CPI data.

We would see the provision of an imputed rentals-based CPI series (released as an additional analytical series) in that light, and repeat our request from 2004 that such an index be created. Further, a cost of living index, if it were to be demanded, should be produced as an analytical series, as the likely inclusion of interest rates in such a series would lead to monetary policy actions having immediate effects on measured inflation that are counterintuitive to the public. For example, inflation would tend to rise immediately when the official cash rate (OCR) is raised, despite the intention to reduce inflationary pressure. Such effects might perversely affect inflation expectations formation.

Likewise, if a seasonally adjusted series were to be produced, it should also be released as an analytical series. We see the production of seasonally adjusted CPI as a low priority, as users such as ourselves who may like to use seasonally adjusted data are able to adjust it ourselves. The main advantages (that we would see as small) arising from Statistics New Zealand producing such a series is the greater perceived impartiality that would come with it and the consistency of adjustment with other official data.

We support the inclusion of overseas online stores for the collection of prices as we are interested in the rate of inflation experienced by households. Purchases made from overseas stores via the internet are an increasingly important factor in the prices of goods and services acquired by New Zealand households. We also believe further work investigating the importance of this channel has merit.

Further improvement to the scope of the CPI could be made through including in the sample frame people living in rest-homes. This is an increasingly large section of population and we believe their purchasing decisions should be reflected in the CPI.

In general, we support any methodological advances that deliver quality improvements to the CPI.

We also note that for any changes to CPI, or publication of new analytical series, lengthy backdated time series are helpful for research purposes, even if they are estimates.

Frequency of the CPI

We appreciate the costings provided by Statistics New Zealand for the publication of a monthly CPI. Our main focus is always on the maintenance of a high quality and credible CPI. New Zealand remains one of the few industrialised countries not to produce its CPI on a monthly basis and is one regard in which New Zealand does not currently comply with the International Monetary Fund's (IMF) special data dissemination standards (SDDS).

To better comply with international standards, we would like to see a monthly CPI produced. We believe the increasing ability to collect data electronically and spread the survey out across the quarter makes the option of spread pricing a cost-effective means to increase the frequency of publication.

A monthly CPI would allow us to incorporate more timely information into the monetary policy process. Further, it would enhance research into areas where international comparability is important, like the real exchange rate.

Incorporating retail transaction data into the production of the CPI

We support greater use of scanner data in the production of the CPI. The potential to increase the accuracy in relation to quality and quantity adjustment especially in the area of consumer electronics is very appealing. The fast-paced nature of change and churn in these product categories means the increased ability to apply quality adjustment techniques will improve the CPI.

The ability of scanner data to provide extensive information on purchases would lead to better product substitution within the index between reweights, especially with rapidly evolving products. We also believe the current three years between reweighting could be maintained with the knowledge that the index is maintaining relevance more effectively through the use of scanner data.

In our view, the cost of having to employ different indexing methods is far outweighed by the benefits that would accrue to the quality of the CPI.

Regional/subnational data

Previously, we have not recommended resources be applied to developing regional CPI data or indexes. However, the earthquakes in Canterbury and recent price pressures related to housing in Auckland have brought the idea of regional price pressure into focus. The 2004 Advisory Committee recommended moving to regional expenditure weights that would have the added benefit of enabling derivation of regional indexes. While there was no evidence the change would improve the quality of CPI, the ability to derive the regional indexes merits rethinking given current interest in regional data.

Dissemination of CPI information

We support Statistics New Zealand’s initiatives to increase awareness of CPI data and how it can be used, such as by the provision of a personal inflation calculator on the Statistics New Zealand website.