CPI International Comparisons

The Consumer Price Index (CPI) is produced by National Statistical Offices across the world and is generally recognised as the principal measure of inflation experienced by households.

The CPI is routinely used to make comparisons of inflation in different countries. While using the CPI for this purpose is reasonable, there are some important differences in the way the CPI is measured in each country to be aware of when making these comparisons. Differences in approaches include how housing is measured, geographical coverage of the price collection and frequency of CPI basket updates, among other things.

This article discusses key differences in how Australia and selected countries, namely New Zealand (NZ), the United States (US), the United Kingdom (UK) and Canada, as well as the European Union (EU), measure CPI inflation. How the CPI is measured in each country differs based on geographical and population coverage, product (goods and services) coverage, methods, data sources, and importantly, the principal purpose of the CPI. Many countries produce the CPI for the dual purpose of a macro-economic measure of inflation to inform monetary policy where inflation targeting is used, and for indexation purposes, where pensions and other government payments are increased in line with CPI inflation.

This article provides a comparison of recent inflationary trends and discusses the different methods applied in each country. A focus of the article is whether owner-occupied housing costs are included in the CPI for each country, along with the reason behind these differences.

Prior to the COVID-19 pandemic, annual inflation in the countries being compared was relatively low and stable, generally tracking below 3 per cent. From 2021, the increase in inflation shown in graph 1 represents the highest inflation experienced over the past 30 years.

Among the selected countries, CPI inflation started to pick up first in the US in early 2021. Higher inflation followed in Australia, New Zealand, Canada and the UK over the next six months. While the increase in inflation in Australia may have started later, and been slower to rise, Australia has caught up to other countries. In Australia, inflation peaked at the end of 2022 at 7.8 per cent, which is the highest in 32 years.

The method used to measure owner occupied housing (OOH) is an important factor in how the CPI is measured. Importantly, the weight for OOH is the biggest component in the selected countries’ CPI (except the EU). Defining OOH in the context of CPI measurement is not straight forward and a lack of reliable data remains a challenge for many countries. There are generally four common approaches to measuring OOH in the CPI, including those used by comparable countries.

1. Rental equivalence: this approach measures the notional price that an owner occupier would pay if they were renting their own dwelling. This uses the rent paid for an equivalent dwelling in the private sector as a proxy for the costs faced by an owner-occupier.
2. Net acquisitions: this approach treats a house as the purchase of a good that is part asset (the land) and part consumable (the structure) with the price of land being excluded from the CPI. As a result, this method includes net purchases of dwellings by the household sector (excluding land prices), alterations and additions, transaction costs, such as tax and legal fees, and running costs, such as repairs, maintenance, and insurance.
3. User cost: this is a use-based approach. It treats housing as a capital good that provides services to measure the costs of owning the house. It includes actual expenses incurred such as repairs, maintenance and the costs of financing the purchase of the property.
4. Payments (or Outlays) approach: defined by looking at what households pay out as owner occupiers (excluding capital payments). This includes mortgage interest component payments, transaction costs (stamp duty, estate agency fees and conveyancing) and running costs (such as repairs, maintenance, and insurance).

In selecting the most appropriate approach to measuring OOH, the CPI International Manual states:

Ideally, the approach chosen should align with the conceptual basis that best satisfies the main use of the CPI. However, the data requirements may be such that it is not feasible to adopt the preferred treatment. Also, the dual use of CPIs as both macroeconomic indicators and for indexation purposes can lead to clear tensions in designing an appropriate treatment for owner-occupied housing services costs that suits all needs.¹

Table 1 shows which approach to owner-occupied housing measurement are used by comparable countries.

The four approaches can result in quite different weights in the CPI basket and contributions to CPI inflation. For example, the Net acquisitions approach results in a lower weight for OOH compared to the Rental equivalence approach.

Australia and New Zealand use the Net acquisitions approach which aligns to the principal purpose of the CPI as a macro-economic measure of inflation. Data is also available separately on the price of land and the structure, which enables the land component to be excluded due to it being considered an asset rather than consumption. The ABS also publishes the Selected Living Cost Indexes² for different households and use the Payments approach to measure OOH.

Rental equivalence is used by the UK and US, while Canada has adopted a User cost approach. The EU does not currently capture OOH due to data limitations in many member countries, but there is ongoing work to consider whether a Net acquisitions approach is possible.

The advantage of the Net acquisitions approach is that it attempts to measure the transaction price of the dwelling, where the price reflects changes in the cost of materials and labour. This is consistent with the approach used for other durable goods such as motor vehicles and furniture. A challenge some countries face with this approach is separating the value of the structure and the land, with the land being considered an asset and out of scope of the CPI. The ABS excludes the value of land by collecting prices on project homes for houses and measuring the cost of construction for apartments. A criticism of the Net acquisitions approach is that, by excluding the cost of the land, it’s not capturing the experience of households, where land is an inescapable cost of owning your own dwelling.

The Rental equivalence approach is simple to apply in practice, however, it has the disadvantage of being an imputed price. With the concept being used of an owner occupier renting the dwelling to themselves, there is no actual transaction occurring, so an imputed price is needed to measure price change. The rental equivalence approach simply values the services provided from owning a dwelling by the corresponding market rental value of a similar dwelling.

The Payments approach, and to a lessor extent, the User cost approach, are largely impacted by changes in mortgage interest rates. This is how a majority of households purchase their dwelling and is an intuitive way to think about the cost of owning a dwelling. A disadvantage is that for countries that have inflation targeting, an increase in interest rates has a direct upward impact on the CPI, when the purpose of higher interest rates is to lower inflation. This circular feature makes it difficult for central banks to use the CPI as their inflation target when the Payments approach is used to measure OOH.

Graph 2 shows the annual percentage change in the OOH component over the past five years. Prior to 2021, the price change of OOH in each country was similar and relatively low and stable. Over the past two years, there has been a significant divergence in measured OOH and its contribution to CPI inflation. In the case of Australia and NZ, OOH has been a significant contributor to CPI inflation with an annual movement of around 20 per cent for the OOH series, which was over twice the rate of CPI inflation. While in the UK and US, the increase in OOH has been much smaller at around 3 per cent – less than half of the rate of CPI inflation.

In the case of Australia and NZ, price increases for OOH had an upward contribution on CPI inflation, while in the US, Canada and the UK it had a moderating impact on the CPI in these countries.

The increase in OOH in Australia and NZ was due to higher material costs and strong demand for new dwelling construction. Due to the different methods used to measure OOH, this was not captured in the CPI of the other countries, despite higher global prices for timber, steel, concrete and other construction materials.

This demonstrates that while attempting to measure OOH, the different approaches can result in very different contributions to CPI inflation.

With Australia’s use of the Net acquisitions approach to measuring OOH, it aligns closely to the cost of dwelling construction, which includes the cost of both materials and labour. This can be seen in graph 3 in how the CPI New dwellings series is similar to the Input to house construction series measured in the Producer Price Index.

If Australia were to use the Rental equivalence approach, the increase in OOH would have looked closer to the CPI Rents series, whereas a Payments or User cost approach would look similar to the Mortgage interest charges series used in the ABS’s Selected Living Cost Indexes.

The use of the Rental equivalence approach would have lowered CPI inflation in 2021 and 2022 due to rent prices increasing by less than dwelling construction costs. The Payments approach and the inclusion of mortgage interest charges would have lowered CPI inflation up until June 2022 quarter due to lower interest rates and then increased CPI inflation from September 2022 quarter onwards as interest rates started to rise.

When it comes to second hand goods, such as used motor vehicles, transfers between households are not included in the CPI as one household’s spending is another household’s income, so the net effect of expenditure is zero. For used motor vehicles, what is in scope are transfers between governments and businesses to households. In addition to this, the margin a dealership charges when a used vehicle is purchased from one household and sold to another household. In practice, it is difficult to price used motor vehicles to a consistent level of quality from one quarter to the next where the condition of the vehicle and number of kilometres travelled impacts the price of the used motor vehicle. It is also challenging to measure the dealership margin in practice for transfers between households. Therefore, in Australia’s CPI, used motor vehicles are excluded with only new motor vehicle purchases measured.

Some countries do include used motor vehicles in their CPI, such as the US. Graph 4 shows that in the US price change is usually similar for new and used motor vehicles. However, following the COVID-19 pandemic, a shortage of new motor vehicles resulted in a significant increase in demand for used motor vehicles. In 2021, new motor vehicle prices in Australia increased by around 6 per cent and in the US it was around 10 per cent, while used motor vehicles in the US increased by as much as 45 per cent in the 12 months to June 2021.

While some countries do include used motor vehicles in their CPI, this is not the case for other durable goods that can be purchased second hand, such as furniture and whitegoods. This leads to an inconsistency in how some second hand goods are included in the CPI, while others are not.

The way the CPI is measured by countries can differ in many other respects. Some other notable differences are how often the weights of the CPI basket are updated and the data sources used to measure the CPI.

Each year the ABS updates the weights used for the CPI basket³. This has been important for capturing changes to spending patterns in recent years, such as the fall in spending on holiday travel when borders were closed and subsequent increase when travel restrictions were lifted. CPI weights are also updated each year in the UK, Canada, and the EU, while in the US and NZ it’s every 2-3 years.

Big data, such as scanner data and web-scraped data, present opportunities to measure the CPI more accurately, however, its use across countries differs substantially. In Australia, the ABS has continued to increase the use of innovative data sources. In particular the use of weekly scanner data is used to measure price change for supermarket grocery items, which make up around 16 per cent of the CPI basket.

In the case of the weekly scanner data, the ABS was able to dynamically measure the CPI to reflect the change in spending patterns since COVID-19. This was particularly important during the early stages of the pandemic when there was significant stockpiling by households and supply shortages at supermarkets. The scanner data showed that spending on non-perishable goods increased significantly more than spending on perishable goods. For example, the amount spent on toilet paper doubled in March and April 2020. Without access to weekly scanner data, the ABS would not have been able to capture the larger increase in demand for non-perishable goods. This meant the ABS was able to use the scanner data to re-weight the part of the index measuring grocery products in real time to more accurately reflect the impact that any changes in price for these items would have had on the overall CPI.

1. Paragraph 11.83, Consumer Price Index Manual: Concepts and Methods, 2020
2. Living cost indexes can be used to assess changes in the purchasing power of the disposable incomes of households. It measures the impact of changes in prices on the out-of-pocket expenses incurred by households to gain access to a fixed basket of consumer goods and services.
3. Latest link to annual weight update of the CPI and Living Cost Indexes
4. Sourced from Retail Trade, Australia, August 2020
5. Sourced from Consumer price indices (CPIs) - Complete database (oecd.org)