6401.0 - Consumer Price Index, Australia, Sep 2013 Quality Declaration 
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 23/10/2013   
   Page tools: Print Print Page Print all pages in this productPrint All

FEATURE ARTICLE THE USE OF TRANSACTIONS DATA TO COMPILE THE AUSTRALIAN CONSUMER PRICE INDEX


INTRODUCTION

The purpose of this article is to advise the user community of the significant increase in use of transactions data to compile the Consumer Price Index (CPI). This will occur from the March quarter 2014.

Transactions data provide the ABS with opportunities to improve the quality of statistics produced for the Australian community. Transactions data contain detailed information about the business name and location of the transaction, date and time, quantities, product descriptions, values of products sold as well as their prices.

The use of transactions data to compile key economic statistics such as the CPI has been under investigation internationally for some time. Recently the ABS obtained transactions data from a selection of Australian businesses. These data have been used to assess methods and develop processes to compile the CPI using transactions data. The ABS has now sufficiently resolved the various challenges to be able to commence a phased implementation of transactions data in the Australian CPI.


THE CURRENT CPI DATA SOURCES AND SAMPLING APPROACH

Most prices currently used to compile the CPI are collected by personal visits to selected businesses. These personal visits are made by ABS field officers who observe prices as well as discuss discounts, special offers and volume-selling items with the businesses. The field officers record this information during the visit in handheld computers. The regular personal visits to businesses enable the ABS field officers to actively monitor market developments and observe product quality change.

Prices for a small number of products are currently obtained from transactions data. These prices are used to compile the CPI. Automotive fuel prices, for example, are obtained for a sample of electronic funds transfer transactions in each capital city. Prices from businesses across all areas of each capital city are obtained each day, including weekends and public holidays. Prices are recorded for a range of automotive fuel types.

The ABS uses non-probability sampling to compile the CPI. This sampling approach selects representative sets of products for regular pricing from a selection of businesses. The price of every variety of good and service purchased by the consumer is needed to construct a perfectly accurate CPI. This would mean collecting a full set of prices from every outlet. As this is not feasible in practice, most prices are sampled from a selection of outlets in a sample of locations chosen to be representative of the CPI population.


TRANSACTIONS DATA

Definition

The 2008 System of National Accounts defines a transaction as “an economic flow that is an interaction between institutional units by mutual agreement.”(footnote 1) Transactions data are a description of the interactions of institutional units buying and selling products on terms mutually agreed by the buyer and seller.

Transactions data have a number of characteristics. The data provide detailed information about the business name and location of the transaction, date and time, quantities, product descriptions, values of products sold as well as their prices.

In the case of retail outlets, transactions data are often obtained by ‘scanning’ the bar codes for individual products at electronic points of sale.


The benefits of using transactions data to compile the CPI

The ABS initiative to include transactions data in the CPI is driven by three key reasons. They are:

      i. To improve the accuracy of the CPI;
      ii. To reduce data collection costs; and
      iii. To position the ABS to expand the range of consumer price measures.

i. Improving the accuracy of the CPI

The accuracy of the CPI is improved when transactions data are used to:
  • Increase the frequency of price observations. Average transaction prices experienced by consumers over a period such as a week or a month can be calculated;
  • Increase product and business coverage. Transactions data enables the ABS to cost effectively increase the coverage of products sampled.
  • Inform sampling decisions. Transactions data provide revenue and quantity information. This provides valuable information to identify business outlets to include in CPI samples.
  • More frequently update weighting information. This is obtained from quantity information contained in the transactions dataset.

ii. Reducing data collection costs

Collecting ‘shelf’ prices by personal visit to businesses is a significant cost in producing the Australian CPI. Obtaining a consistent and timely supply of transactions data from businesses would reduce the work involved in data collection.

iii. To position the ABS to expand the range of consumer price measures

The ABS, subject to available funding, will examine transactions data to determine whether additional consumer price measures can be produced from transactions data. For example, regional and monthly temporal price indexes may be able to be produced from the data for selected product groups. Spatial price indexes may also be able to be produced.


Using transactions data in the CPI

International experience

Using transactions data to compile the CPI has received considerable attention from price statisticians internationally. Typical of very large datasets, there are challenges associated with their use. Various methods have been examined internationally over the previous decade to determine how best to utilise such a large volume of very detailed, high frequency data to compile the CPI. Most recently the ‘Rolling Window GEKS’ (RWGEKS)(footnote 2) approach has emerged as the leading methodology.

The RWGEKS method builds on the multilateral method of Gini (1931), Eltetö and Köves (1964) and Szulc (1964) (GEKS), which is known from spatial price comparisons and applies this method to price comparisons across time. The GEKS method takes the geometric mean of the ratios of all bilateral indexes (calculated using the same index number formula) between a number of entities. For spatial indexes these entities are generally countries, while for price comparisons across time, the entities are time periods (usually months).

A problem with this approach is that the results for all time periods will change when the observation period is extended and new data are added. By consequence index numbers will be subject to continuous revision, something which is unacceptable for a CPI. The RWGEKS approach was suggested to address this problem.

The RWGEKS approach uses a moving window to continuously update the observation period as data for new periods become available and calculates the price change between the two most recent periods, without the need to revise previous periods. This approach is undertaken as the earlier data in the sample become less and less relevant for later comparisons.

Work is still continuing to resolve some RWGEKS methodological challenges, including quality change associated with disappearing products and their replacements. At this time no single method to incorporate transactions data into the CPI has received international endorsement.

A small number of National Statistics Offices (NSOs) have utilised transactions data to compile their CPI. The use of transactions data varies from replacement of field collected prices(footnote 3) to implementation of new index construction methods(footnote 4) .

The ABS will continue to engage with the international community to contribute to the development of an internationally endorsed methodology to use a large volume of high frequency data to compile the CPI.

The ABS approach

The ABS will implement transactions data in the Australian CPI in two phases.

The first phase will commence on 1 January 2014. The ABS will replace field collected prices with prices derived from transactions data. ABS field officers will no longer personally visit businesses that have provided the ABS with transactions data.

The price for an individual product is calculated from the transactions data by dividing a product’s revenue by the quantity sold. This price is referred to as a product unit value and represents the price experienced by consumers over a period of weeks or months. A product’s unit value is more representative of prices paid by consumers over the reference period than point-in-time pricing.

Larger price samples for individual products will be available from transactions data. A unit value price for individual products will be collected from an increased number of business outlets in each capital city.

Product unit values obtained from transactions data will be used to compile a range of CPI Expenditure Class indexes. These Expenditure Classes are listed in Attachment 1.

In the second phase, the ABS will assess the suitability of an internationally developed transactions data methodology and implement it as appropriate to compile the CPI.

The ABS is currently examining the RWGEKS method. The ABS will produce experimental indexes using this methodology. Results of this work will be published and discussed with the user community.


Future directions

ABS work has also commenced to determine whether transactions data can be used to compile a range of other macroeconomic statistics in the areas of retail activity and household consumption. Results of this work will be published as they become available.


Attachment 1: Expenditure Classes where some prices will be derived from transactions data
  • Bread
  • Cakes and biscuits
  • Breakfast cereals
  • Other cereal products
  • Beef and veal
  • Pork
  • Lamb and goat
  • Poultry
  • Other meats
  • Fish and other seafood
  • Milk
  • Cheese
  • Ice cream and other dairy products
  • Fruit
  • Vegetables
  • Eggs
  • Jams, honey and spreads
  • Food additives and condiments
  • Oils and fats
  • Snacks and confectionery
  • Other food products n.e.c.
  • Coffee, tea and cocoa
  • Waters, soft drinks and juices
  • Take away and fast foods
  • Tobacco
  • Garments for men
  • Garments for women
  • Garments for infants and children
  • Glassware, tableware and household utensils
  • Tools and equipment for house and garden
  • Cleaning and maintenance products
  • Personal care products
  • Pharmaceutical products
  • Medical and hospital services
  • Spare parts and accessories for motor vehicles
  • Automotive fuel
  • Newspapers, magazines and stationery
  • Pets and related products
  • Other financial services
  • Property rates and charges


References

Eltetö, O. and P. Köves (1964) 'On a problem of index number computation relating to international comparison', Statisztikai Szemle, 42, pp. 507-18.

Gini C. (1931) 'On the circular test of index numbers', International Review of Statistics, Vol. 9.

Ivancic, L., W.E. Diewert and K.J. Fox (2011), Scanner Data, Time Aggregation and the Construction of Price Indexes, Journal of Econometrics, Vol. 161, Issue 1, 24-35.

Sammar, M, Norberg, A. and Tongur, C. (2013), ’Issues on the use of scanner data in the CPI’, Paper presented to the Thirteenth Meeting of the International Working Group on Price Indices, Copenhagen, Denmark.

Szulc B. (1964), 'Indices for multiregional comparisons', Przeglad Statystyezny 3, Statistical Review 3, pp. 239-54.

United Nations, International Monetary Fund, Organisation for Economic Co-operation and Development, World Bank and Commission of the European Communities, Brussels/Luxembourg, New York, Paris, Washington D.C., 2008 System of National Accounts.

van der Grient, H.A. and J. de Haan (2011), Scanner Data Price Indexes: The ’Dutch Method’ versus RYGEKS, Paper presented at the Twelfth Meeting of the International Working Group on Price Indices, Wellington, New Zealand.

Footnotes:
1 2008 System of National Accounts, p39, [3.7] <back
2 Ivancic, L., W.E. Diewert and K.J. Fox (2011), Scanner Data, Time Aggregation and the Construction of Price Indexes, Journal of Econometrics, Vol. 161, Issue 1, 24-35 <back
3 Sammar, M, Norberg, A. and Tongur, C. (2013),’ Issues on the use of scanner data in the CPI’, Paper presented to the Thirteenth Meeting of the International Working Group on Price Indices, Copenhagen, Denmark <back
4 van der Grient, H.A. and J. de Haan (2011), Scanner Data Price Indexes: The ’Dutch Method’ versus RYGEKS, Paper presented at the Twelfth Meeting of the International Working Group on Price Indices, Wellington, New Zealand. <back