6401.0 - Consumer Price Index, Australia, Sep 2018 Quality Declaration 
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 31/10/2018   
   Page tools: Print Print Page Print all pages in this productPrint All

70 years of the Australian Consumer Price Index

The Consumer Price Index (CPI) is a statistic that impacts the lives of all Australians. The principal purpose of the CPI is a macro-economic indicator of general consumer price inflation in the Australian economy. The CPI is widely used by governments, economists and academics for macro-economic analysis, as well as by businesses and the general community in determining employee wages and the prices of goods and services paid by households. It is also the principal measure of inflation used by the Reserve Bank of Australia (RBA) in setting monetary policy.

While the Australian Bureau of Statistics (ABS) has measured inflation in Australia for more than 100 years, this September quarter marks 70 years of producing the CPI in its current format. This makes the CPI the longest continuous time series published by the ABS

This paper will give a brief history of the ABS's measurement of inflation and discusses how the methods used by the ABS have evolved over time. While the CPI continues to be a robust measure of consumer price inflation, with new data sources becoming available, and innovative methods being developed, this paper will also highlight how the ABS has enhanced the measurement of the CPI in recent times.


HISTORY OF THE CPI

1910 - Creation of the Retail Price Index

The ABS was formed following the creation of the Census and Statistics Act 1905. The first Commonwealth Statistician, George Knibbs, soon realised the importance of measuring price change in the Australian economy. In 1910 Knibbs began work on Australia's first Retail Price Index. Knibbs noted the pressing need for this work:

"Probably no subject has recently attracted a greater amount of public attention than the extent of variations in prices, a matter which, in its relation to the cost of living, is fraught with importance to every section of the community." (Price Indexes and the Cost of Living in Australia, Labour and Industrial Branch Report, 1912)

Following an enquiry into the cost of living, the ABS published in 1912 the first report on 'Prices, Price Indexes and Cost of Living in Australia'. The report contained the measurement of price change for food, clothing and rent for the six capital cities and for 30 towns. This information continued to be released on an annual basis, with a lag of around one year.

1920s - 1950s

Over the next forty years the Retail Price Index went through various changes to include new products into the index. Before the introduction of the CPI in 1960, there were a number of retail price indexes compiled by the then Commonwealth Bureau of Census and Statistics.

The 'A-series' was the first and saw credit with its adoption by the Federal Arbitration Court to determine relative basic wage in Australia. In 1925, the 'B-series', replaced the A-series for general statistical purposes, however it was not used in wage determination. These two series' were discontinued in 1938 and 1953, respectively.

By far the most important of these initial price indexes was the 'C-series Index', which was the principal retail price index in Australia for almost forty years. It was first compiled in 1921 with index numbers compiled back to 1914. The C-series Index was reviewed in 1936 and a slightly revised selection of goods and services was introduced, which then remained unchanged until the C-series Index was discontinued in 1960.

One of the most difficult challenges for the measurement of inflation was following the outbreak of war in 1939. This period saw policy changes in the form of various wartime controls, such as food rationing, causing recurrent changes in household consumption patterns. In 1953, the decision was made to continue compiling the C-series Index on its pre-war basis, but also to compile an interim retail price index based on the post-war pattern of household expenditure.

The Interim Retail Price Index replaced the C-series in 1954 based on post-war consumption patterns and included a greater amount of items not covered by the preceding series'. The C-series was kept for its pre-war basis, still being compiled during this time and continued to be the official index while the Interim Retail Price Index was used for general statistical purposes.

1960s - 1990s

In 1960 there was a significant change in the approach to the measurement of retail price change. Instead of using long-term fix weighted indexes, the approach was changed to produce a series of short-term indexes that would then be linked to form a long-term series. This can be considered the birth of the modern CPI. This approach, still being used today, produces short-term indexes for approximately five years, which are then linked to a long-term series.

The motivation for this change was outlined in 'A Description of the Consumer Price Index', ABS, 1960:

It was clear that no single list of items and no single set of fixed weights would be adequately representative as a basis for measuring retail price changes at all times throughout the post-war period. In consequence, the situation was met by compiling the Consumer Price Index constructed as a chain of linked indexes with significant changes in composition and weighting effected at short intervals.

The CPI was first compiled in 1960 with index numbers compiled back to the September quarter 1948. The CPI was designed to measure quarterly changes in the retail prices of goods and services purchased by metropolitan wage earning households.

In the mid-1970s, the ABS conducted the first Household Expenditure Survey (HES). The data from the HES was used in the CPI for weighting purposes, which helped ensure the CPI basket reflected contemporary household expenditure. The HES continues to be an important data source for updating the CPI.

For the next 20 years the methods used in the CPI were largely unchanged, with the exception of 'Mortgage interest charges' replacing the expenditure class 'House purchase' in 1984.

1990s

In 1997 a major review was conducted on the CPI, which led to a re-think of the scope and purpose of the CPI. The review saw the following changes introduced in 1998:

  1. The primary purpose of the CPI changed from a measure of living costs of employee households to a general measure of price inflation for the household sector as a whole.
  2. The coverage of the CPI weights was expanded to all private households in the eight capital cities; prior to this the coverage was only metropolitan wage-earning households. This saw the population coverage increase from 29 per cent of Australian private households, to 64 per cent.
  3. Mortgage interest charges were removed from the CPI and replaced with expenditure on new dwellings (excluding the land).

The 1990s also saw the Reserve Bank of Australia (RBA) introduce an inflation target of between 2 and 3 per cent annual change in the CPI. This cemented the CPI as the primary indicator to informing monetary policy and the setting of interest rates.

2000s - 2018

In the 2000s the biggest change to the CPI was the introduction of the measurement of financial intermediation services indirectly measured (FISIM). This saw Australia as the only country in the world to include the measurement of FISIM in its CPI. Following a review of the CPI in 2010, FISIM was removed in 2011. Also in 2011, the ABS first published seasonally adjusted measures for the CPI and two underlying inflation series, the 'Trimmed mean' and 'Weighted median'.

In recent years the CPI has continued to evolve, in particular, seeing greater use of administrative data, such as retail transactions 'scanner' data, and web-scraped data. These new data sources have reduced the collection costs and enhanced the accuracy of the CPI. In 2017 the ABS has adopted an innovative method, known as a multilateral index method, to make greater use of the price and expenditure information contained in the transactions data. The ABS is leading the world in the use of these methods.

A significant change occurring in the December quarter 2018 is the move to annually re-weighting the CPI using data from the National Accounts. While the HES data will continue to be an important data source when it's available, National Accounts data will be used to update the weights for the years in-between the availability of data from the HES. This change will see more up-to-date weights being used, further enhancing the accuracy of the CPI and bringing the ABS into line with international best practice. These changes are discussed in more detail in the following section.

CHART 1: ANNUAL CPI MOVEMENT (%)
Chart 1: Annual CPI movement (%)



INNOVATION IN THE CPI

The ABS has a long history of adopting innovative methods and new data sources to compile our statistics. In the case of the CPI, the methods and data sources have continually evolved in order to align to international standards and best practice. In particular, over the past five years, the ABS has used innovation to enhance the quality of the CPI and, at the same time, reduce the cost of producing the CPI.

Incorporating new data sources into the CPI

Price collection in the CPI has traditionally involved an ABS field officer obtaining price and product information by visiting retail outlets in person or via telephone. This is considered a relatively costly and inefficient approach when compared to other data sources, such as administrative data. Administrative data is information collected for other purposes by governments and businesses that can be used for the compilation of official statistics. In recent years, the ABS has been making more and more use of administrative data to produce the CPI.

The most significant progress in the use of administrative data in the CPI has been the introduction of transactions 'scanner' data in 2014. Scanner data refers to the transactions that are recorded (or ‘scanned’) at the retailer’s point of sale. Scanner data is high in volume and contains detailed information about individual transactions including: date of purchase, quantities purchased, product descriptions, and value of products purchased. The use of scanner data has replaced approximately 25 per cent of the CPI's manual price collection.

In addition to administrative data, the ABS has also been gradually introducing online data collected through the use of ‘web-scraping’ processes to replace manual price collection across a number of CPI products. Web-scraping makes use of automated software to extract large amounts of prices and product information from websites. Prices can be collected as frequently as desired for all products using purpose-built programs that scan the websites of retailers to find the relevant information. Web-scraping has enabled approximately 500,000 prices to be collected per week compared to 1,000 prices with traditional CPI collection, providing an opportunity to significantly increase the amount of products sampled.

Innovative methods used in the CPI

With the availability of scanner containing both price and expenditure information, in 2017, the ABS introduced the use of an innovative method to make greater use of the scanner data in the CPI. This new method, known as a 'multilateral index method', utilises a census of products available in the scanner data, uses the expenditure data to weight the products each quarter, and reduces data collection and processing costs. The ABS is leading the world in the use of multilateral index methods in the CPI. For more information see the information paper An Implementation Plan to Maximise the Use of Transactions Data in the CPI (cat. no. 6401.0.60.004).

In 2018, the ABS, for the first time, will annually re-weight the CPI using Household Final Consumption Expenditure (HFCE) data from the National Accounts to derive expenditure weights for the CPI basket for the years between the release of the ABS's Household Expenditure Survey, currently available every six years. ABS research has shown that annual re-weighting will improve the accuracy of the CPI by better capturing changes in household spending patterns. For more information see the information paper An Implementation Plan to Annually Re-weight the Australian CPI (cat. no. 6401.0.60.005).

Future innovation

With the greater use of administrative data and the introduction of automated data collection and processing techniques, this has led to both an increase in the quality of the CPI and a decrease in the cost of producing the CPI. These innovations have positioned the ABS to produce the CPI on a more frequent basis, from quarterly to monthly. Australia is one of only two OECD countries that do not produce their CPI on a monthly basis, the other country being New Zealand. There has also been continued demand for a monthly CPI from policy makers and users, such as the Reserve Bank of Australia, where a more frequent CPI would provide an earlier indication of the trend of inflation and identifying turning points in the economy.

In 2018, the ABS began development work on the feasibility of producing a monthly CPI. An outcome from this work is expected in 2019. For more information see Towards an Australian Monthly CPI.


CHANGES TO THE CPI BASKET

The CPI basket undergoes regular updates to reflect the changing consumption patterns of Australian households. The make-up of the CPI basket has two aspects: the weights contained in the basket; and the goods and services actually included in the basket. Both aspects need to be updated regularly to ensure the CPI remains representative of the average expenditure of Australian households. Here we look back at how the CPI basket has evolved overtime.

CPI weights

The CPI is compiled at various levels of product categories. The structure of the CPI starts with the prices paid for the goods and services that are purchased by households. These prices are then allocated to a category such as 'Bread, pre-packaged, white'. The CPI is then compiled for broader and broader product categories, such as 'Bread', then 'Bread and cereal products', then 'Food and non-alcoholic beverages', and finally, the overall CPI. At each level of these categories an expenditure weight is assigned to reflect the contribution it makes to the CPI. For example, a 10 per cent rise in the price of milk, with a weight of 0.5 per cent in the CPI, will have less of an impact on the CPI than a 10 per cent rise in the price of automotive fuel, with a weight of 3.3 per cent in the CPI.

Looking at how the weights in the CPI have changed overtime provides an insight into how the consumption patterns of households have evolved. Table 1 shows some interesting changes to the weights in the CPI basket.

The most significant change has been for Food and non-alcoholic beverages with its weight halving from 32 per cent in 1960 to 16 per cent in 2018. The fall in weight for Clothing and footwear, 17 per cent to 3.5 percent, and Furnishings, household equipment and services, 12 per cent to 9.5 per cent, partly reflects how the prices of imported goods have become cheaper in Australia.

Categories that have increased in weight show that as household disposable income has grown Australians spend more on Housing, 15 per cent to 23 percent, and Recreation and culture, 3 per cent to 13 per cent. Also, with the ageing population, the weight for Health has increased from 1 per cent to 5.5 per cent.
TABLE 1: CPI GROUP WEIGHTS (%)

1960
1980
2000
2018

Food and non-alcoholic beverages
32.1
21.0
17.7
16.1
Alcohol and tobacco
8.0
10.2
7.4
7.1
Clothing and footwear
16.9
10.4
5.1
3.5
Housing
14.9
15.8
22.2
22.7
Furnishings, household equipment and services
12.0
11.5
8.2
9.4
Health
0.8
2.0
4.7
5.4
Transport
11.3
16.6
15.3
10.3
Communication
0.8
1.5
2.9
2.7
Recreation and culture
3.2
8.6
12.3
12.7
Education
n.a.
n.a.
2.7
4.3
Insurance and financial services
n.a.
2.4
1.5
5.8

n.a. = 'not applicable'. These categories were not included in the CPI at the time.


CPI goods and services

When the ABS first measured consumer inflation in 1912, only three product categories were included in the CPI basket: food, clothing and rent. In all, there were 46 items priced in the CPI, among these were:

bread, flour, tea, coffee, sugar, rice, sago, jam, oatmeal, raisins, currants, starch, soap, potatoes, onions, kerosene, milk, butter, cheese, eggs, bacon, ham, beef, mutton, pork.

Table 2 is a summary of the changes to the CPI basket over the years:
TABLE 2: GOODS AND SERVICES ADDED TO THE CPI BASKET

1930sself-raising flour, golden syrup, dried apricots, canned fruit, tinned salmon, newspapers, light globes, electric irons, rents.
1960ssoft drink, chocolate, ice-cream, spaghetti, toothpaste, baby food, cough mixture, electric toaster, refrigerator, washing machine, lawn mower, motor vehicles, hairdressing, dry cleaning, cinema tickets, televisions, house purchase.
1970stakeaway food, restaurant meals, wine and spirits, holidays, fish, insurance, books, toys, sports equipment.
1980smortgage interest charges, consumer credit charges, education, child care, glasses, watches, veterinary fees.
1990shouse purchase, water and sewerage, home computers and software, domestic services (house cleaning, gardening, etc.), tertiary education fees.
2000sfinancial services.


The CPI basket has continued to evolve in order to reflect the new goods and services purchased by Australian households. In recent years, the ABS has added products such as smart phones, electronic books, tablets, digital music and games, streaming services, ride sharing and shared accommodation.

Equally important is the removal of goods and services from the CPI basket that are either obsolete or no longer representative of average household expenditure. Some examples include mutton, kerosene, DVD hiring, cassette tapes, VCRs, comic books and international calling cards.

Through the use of the web-scraping collection technique discussed above, the CPI basket now contains close to 100,000 items and uses over 900,000 price quotations each quarter. The CPI basket therefore captures the wide variety of products purchased by Australian households and ensures the CPI continues to be an accurate measure of consumer price inflation.