# Weights and their sources

Latest release
Consumer Price Index: Concepts, Sources and Methods
Reference period
2018

## Introduction

6.1 Weights and their sources describes the procedures that are typically followed in updating the Consumer Price Index (CPI) weights.

6.2 Weighting practices vary at different levels of the CPI. At the published levels, weights are reviewed annually. The Household Expenditure Survey (HES) is used as the primary data source for updating the weights in the years that it is available. For the inter-HES years, Household Final Consumption Expenditure (HFCE) data from the National Accounts is used to re-weight the CPI. At the unpublished levels, the weights can be updated at any time to accurately represent household spending patterns. Below the elementary aggregate, or price-sample level, there are no explicit use of weights.

## Annual re-weighting

6.3 The availability of HFCE data from the National Accounts provides the ABS with an opportunity to update CPI weights more frequently. The ability to re-weight the Australian CPI more frequently has significant benefits to the user community. These include more accurately reflecting consumer spending patterns, addressing stakeholder concerns following the 16th Series CPI review, coherence across macro-economic statistics and improved alignment with international standards.

6.4 The use of expenditure aggregates from the National Accounts is an internationally recommended approach in situations where the time interval between household surveys is large. Within Australia, HFCE data are available on an annual basis as part of the Australian System of National Accounts (cat. no. 5204.0). Practical challenges exist (due to scope and coverage differences) when using HFCE data for CPI weights. These challenges have been examined in detail in the information paper: Increasing the Frequency of CPI Expenditure Class Weight Updates (cat. no. 6401.0.60.002). The plan to re-weight the CPI annually was detailed in the information paper, An Implementation Plan to Annually Re-weight the Australian CPI (cat. no. 6401.0.60.005).

## CPI weights using HES data for the 17th series CPI

6.5 At the level of the index at which the weights are fixed, the ideal is to have a snapshot of all household expenditure. It is important that the data are consistent across the population group, that is, expenditures can be added up without concerns about coverage and double counting, and that all the information is for the same period. For example, if expenditures are for different periods for different items, then these could be affected by changes in economic conditions, tax rates, and population growth. Any of these influences could easily distort the weights. In addition, although an information source might provide an estimate of total sales of an item, it would be necessary to know the proportion of those sales to households as consumers. For example, sales of whitegoods will include sales that are not in scope of the CPI, for example sales to businesses, and to households as owners of rental properties.

6.6 The HES provides the most comprehensive data on household expenditure. The HES is a sample of just under 8,000 metropolitan households. Data are collected using a diary of personal expenditures in which residents aged 15 years and older record their expenditure over a two-week period. An interview questionnaire also collects information about each household's characteristics, expenditures common to all members of the household (e.g. utility bills), and irregular or infrequent expenditures. The latest HES has been collected in respect of 2015-16 and was introduced into the CPI beginning the December quarter 2017.

6.7 Although the HES provides a comprehensive coverage of household expenditures, there are some weaknesses in the HES data for CPI purposes.

6.8 Being a survey, the HES is subject to sampling error. It is possible for the selection of one or several households with exceptionally high expenditure on an item to significantly affect the expenditure estimate for that item in a region, especially in the smaller capital cities.

6.9 Some expenditures recorded in the HES are not fully contemporaneous. Estimates for some items which are more expensive or purchased infrequently are obtained or supplemented by recall, rather than relying on expenditures actually recorded in the diary during the two weeks that the household is included in the survey. Periods over which households are asked to recall expenditure vary depending on the item.

6.10 The HES records all data exactly as reported by the households with no adjustments for known cases of under-reporting, especially expenditure on alcohol and tobacco.

6.11 There are some expenditures required by the CPI which cannot be obtained from households, such as the net insurance charge (gross premiums less claims).

6.12 Thus, various adjustments need to be made to the information reported in the HES before it can be used in the CPI, and for some items the HES data are supplemented or replaced by other data which provide a more accurate estimate of expenditure.

6.13 There are data sources other than the HES that provide estimates of household expenditure on some commodities, but often only nationally. These alternative sources are used for validating the HES data. For most products included in the CPI, these alternative data are similar to the HES estimates. However, historically there have been significant differences between HES and the other data sources for household expenditure on alcohol and tobacco. For both products, households are the major consumers, and the imposition of excise and other taxes on these products means that the alternative estimates are more accurate than the HES estimates which appear to understate household expenditure.

6.14 As a result, HES estimates are adjusted for under-reporting based on factors derived at the national level using National Accounts HFCE data. These national factors are applied to the HES expenditure estimates for each capital city. The 2015-16 HES has estimated expenditure for alcohol at a little under half, and tobacco at a little over one third of the respective National Accounts estimates.

6.15 As households do not always separately identify alcohol and meal expenditures when reporting expenditure on restaurant meals in the HES, a small adjustment is made using the proportion of reported meal expenditure that is alcohol, estimated from an analysis of HES unit records. Meal expenditure is reduced by the proportion of reported meal expenditure that is alcohol, with the difference included in expenditure on alcohol. This adjustment does not impact overall household expenditure.

6.16 Some expenditures are collected in the HES as recalled items rather than as diary entries. The extent of the recall period varies. For some items, it is purchases in the last three months (most whitegoods, furniture and house repairs); last payment for general rates, electricity, and health services; and the last 12 months including motor vehicle purchase, motor vehicle repair and maintenance, education, overseas travel, and house alterations. To the extent that prices for these items change between the time that the household purchased them and their inclusion in the HES, expenditures will not accurately reflect the underlying quantities acquired during 2015-16.

6.17 No adjustments are made for any items where the recall period is the last three months since the adjustments would be small. Most items where the recall period is the last payment are typically those with a short billing cycle, often quarterly, or where there are options to pay periodically (e.g. local government rates). This leaves only a few items such as motor vehicle registration, overseas holiday travel, and education where expenditures could have been incurred up to twelve months before inclusion in the HES. Adjustments are made for these items.

6.18 The adjustment procedure for a recall period of twelve months is as follows. In the first quarter (Q1) of the HES reference year, the period potentially covered will be from the start of the corresponding quarter of the previous year (for those households selected in the first week of the HES reference year) through to the end of the first quarter of the reference year. Similarly, the pricing period for the households first included in the last week of Q1 in the reference year will commence in the last week of Q1 of the previous year. If expenditures are distributed uniformly over the period, then effectively expenditures in the reference quarter and the corresponding quarter of the previous year will only be half that for the intervening quarters. Thus, assigning weights of, say, one for the intervening quarters and 0.5 for the start and end quarters, we can derive a weighting pattern as shown below.

This image illustrates the adjustment procedure for a recall period of twelve months. In the first quarter (Q1) of the HES reference year, the period potentially covered will be from the start of the corresponding quarter of the previous year through to the end of the first quarter of the reference year. The pricing period for the households first included in the last week of Q1 in the reference year will commence in the last week of Q1 of the previous year. If expenditures are distributed uniformly over the period, then expenditures in the reference quarter and the corresponding quarter of the previous year will only be half that for the intervening quarters. Therefore assigning weights of one for the intervening quarters and 0.5 for the start and end quarters, a weighting pattern is derived.

6.19 Thus the adjustment factor for items with a twelve-month recall for the 2015-16 HES is:

$$\large \frac{\mathrm{(I_{S15}+I_{D15}+I_{M16}+I_{J16}) \times 0.25}}{\mathrm{(0.03 \times I_{S14}+0.09 \times I_{D14}+0.16 \times I_{M15}+0.22 \times I_{J15}+0.22 \times I_{S15}+0.16 \times I_{D15}+0.09 \times I_{M16}+0.03 \times I_{J16})}}$$

where $$\mathrm{I}_{S 15}$$ is the CPI index number for the expenditure class for the September quarter 2015 etc.

Using the following hypothetical index numbers:

$$\mathrm{I}_{S 14}=99 ; \mathrm{I}_{D 14}=100 ; \mathrm{I}_{M 15}=101 ; \mathrm{I}_{J 15}=102 ; \mathrm{I}_{S 15}=103 ; \mathrm{I}_{D 15}=104 ; \mathrm{I}_{M 16}=105 ; \mathrm{I}_{J 16}=106$$

The result of the above formula is:

$$\large \frac{(103+104+105+106) \times 0.25}{(0.03 \times 99+0.09 \times 100+0.16 \times 101+0.22 \times 102+0.22 \times 103+0.16 \times 104+0.09 \times 105+0.03 \times 106)}=1.02$$

(which is the adjustment factor to be applied to the recalled price).

6.20 With a generally low rate of price change over 2014-15 and 2015-16, the adjustments made for recall were small. However, these adjustments can have a significant impact during periods of high inflation.

### Salary sacrifice

6.21 Salary sacrifice is an arrangement between an employee and employer whereby part of the employee's pre-tax cash salary is traded for non-cash benefits. Conceptually, these arrangements should be captured in both gross wages and salaries, and household expenditure. In the 2015-16 HES, the salary sacrifice question module collected detailed information on motor vehicles purchased via a salary sacrifice arrangement. In addition, respondents were instructed to exclude expenditure on motor vehicles and related items for those motor vehicles purchased through salary sacrifice from the motor vehicle question modules within the household questionnaire, to ensure no double counting.

6.22 The items that have been adjusted for salary sacrifice include motor vehicles. To ensure the expenditure on motor vehicles includes all motor vehicles purchases (i.e. including via salary sacrifice) some adjustment is needed to the relevant HES items. Therefore salary sacrifice amounts reported against vehicles are allocated to motor vehicle purchases, registration, insurance, motor vehicle repair and servicing, and automotive fuel.

### Outlier and aberrant expenditures

6.23 The HES data were compared across capital cities and over time to validate the 17th series expenditure at the expenditure class level. The HES expenditure in 2009-10 was revalued to 2015-16 dollars to derive the volume changes between the two HES reference years and compared to the 2015-16 HES. Any large differences were then investigated to see if they were valid. For example, there was a large rise in expenditures on rents between the 2009-10 HES and the 2015-16 HES. The increase was due to both a price and volume increase. The volume increase was driven by an increase in the proportion of people renting in Australia's capital cities.

6.24 A number of unit record adjustments were made, in particular to the smaller capital cities where HES sample sizes are smaller and, in general, the standard errors are larger. The outlier adjustment used was winsorisation, which involves replacing an unrepresentative expenditure by the next largest estimate. Where unit record outliers could not be identified, differences were further investigated. A small number of volume changes could not be validated, resulting in adjustments using either alternative volume data or based on market intelligence.

### Expenditures not sourced from HES

6.25 A summary of the CPI expenditure classes where HES data were not used for weighting purposes as part of the 17th series re-weight is provided below. From the December quarter 2018 onwards, the CPI will be re-weighted annually using Household Final Consumption Expenditure (HFCE) data from the National Accounts. For details on the methods to be used under annual re-weighting, see the information paper: Increasing the Frequency of CPI Expenditure Class Weight Updates (cat. no. 6401.0.60.002).

#### New dwelling purchase by owner-occupiers

6.26 New dwelling purchase by owner-occupiers in the CPI includes the 'net additions of household sector dwellings' as a measure of owner-occupier housing costs. This includes new homes (excluding land) and major improvements. Sales of houses that take place between households (generally established dwellings) are excluded so that the weights relate only to net additions to the housing stock arising from household purchases from other sectors (i.e. from businesses such as builders and developers). Expenditure on New dwelling purchase by owner-occupiers comprises of four components: owner-occupied housing, first home owners' grants, alterations and additions, and installed appliances.

6.27 To derive the owner-occupied housing component, expenditure is estimated by multiplying the average value of private dwelling completions for 2015-16 by the change in the owner-occupied housing stock. The average value of private dwelling completions for 2015-16 is obtained from Building Activity, Australia (cat. no. 8752.0). The change in the owner-occupied dwelling stock is sourced from National Accounts HFCE estimates.

6.28 Consistent with standard practice relating to the inclusion of subsidies in the CPI, subsidies paid to first-home buyers are treated as negative expenditure and subtracted from the new dwelling purchase by owner-occupiers house acquisition expenditure.

6.29 Expenditure on alterations and additions is derived from the alterations and additions component of Private gross fixed capital formation (GFCF) from the National Accounts. Expenditure on installed appliances is sourced from HFCE data. Both items are added to the estimate for house acquisition to provide the total expenditure on New dwelling purchase by owner-occupiers.

#### Motor vehicles

6.30 The weight for motor vehicles is derived from National Accounts HFCE data. This weight reflects purchases of new cars, transfer of used cars to the household sector (from business or government) and the service fee from the transfer of second hand cars.

#### Higher Education Loan Program (HELP)

6.31 Expenditure on higher education in the HES includes Higher Education Loan Program (HELP) payments made by households upfront plus any HELP repayments made through the taxation system during the reference period. This measure is not consistent with the concept of an acquisitions based CPI, where expenditures should reflect the cost to households of the education service acquired during the reference period. The CPI scope includes the actual payments made during the period (upfront payments) plus fees for education services acquired during the period but deferred to be paid at a later date.

6.32 To align the household expenditure on tertiary education fees in the CPI, HELP expenditures are calculated using data from the Department of Education and Training on total upfront and deferred fees, and the number of students paying HELP loans.

#### Insurance

6.33 The Insurance EC covers comprehensive insurance for dwellings (including contents), motor vehicles and compulsory third party (CTP) insurance. For the purpose of measuring household price inflation, the weight for insurance in the CPI should reflect the cost of the service provided by insurers (gross premiums less claims), rather than the value of gross premiums paid. The latter is reported in the HES.

6.34 Expenditure on contents, motor vehicle and CTP insurance are derived from National Accounts HFCE data. The insurance for dwellings component is excluded from HFCE as it is considered intermediate consumption in the National Accounts. Expenditure on insurance for dwellings is instead sourced from National Accounts estimates of intermediate consumption of the Ownership of dwellings industry. These National Accounts estimates align with the conceptual basis of Insurance in the CPI.

6.35 Expenditures funded by claims are added back to the appropriate items. This data is also sourced from the National Accounts.

#### Financial services

6.36 For most financial services, expenditure cannot be sourced from the HES as it is either not directly observable or the HES does not capture the transactions in sufficient volumes or detail. The Financial services sub-group includes two expenditure classes: Deposit and loan facilities (direct charges) and Other financial services. Detailed information on Financial services in the 17th series CPI is contained in Appendix 3.

#### Deposit and loan facilities (direct charges)

6.37 Expenditure on Deposit and loan facilities (direct charges) is determined through the use of administrative data sets (obtained from financial institutions and government reporting agencies) of financial institution fees and charges for Australian households.

#### Other financial services

6.38 Other financial services include real estate agent services, legal and conveyancing services, stockbroking services, accounting services and taxes on property transfers (stamp duty).

6.39 The real estate fees component of the National Accounts Private GFCF ownership transfer costs series is used to derive expenditure on real estate agent services.

6.40 Expenditure on legal and conveyancing services is also derived from the Private GFCF ownership transfer costs series from National Accounts.

6.41 National Accounts HFCE data is used to derive an estimate of household expenditure on stockbroking services.

6.42 Expenditure on accounting services is derived from HES data.

6.43 Taxes on property transfers is compiled using estimates from Taxation Revenue, Australia (cat. no. 5506.0) and data supplied by the State and Territory Revenue Offices.

6.44 The ABS will investigate methodologies for other significant financial services that are currently not covered in the CPI (e.g. superannuation charges) and introduce them into the CPI when the ABS is satisfied that the methodology and data are sufficiently robust to produce high quality estimates.

#### Revaluing expenditures to the link period

6.45 The expenditure weights derived from HES are based on expenditures (i.e. price x quantity) in 2015-16 (the weight reference period). This new expenditure pattern was not introduced into the CPI until the December quarter 2017 (with September quarter 2017 as the link period). Prior to implementation, in line with current CPI and international practice, the expenditures have been revalued to the September quarter 2017 to preserve the underlying quantities, but take into account the price changes that have occurred between the weight reference and link periods.

6.46 The calculation of the revalued estimates involves taking the 2015-16 expenditures and multiplying them by revaluation factors. These factors are derived as the ratio of the component's September quarter 2017 price index to the average of its quarterly price indexes for 2015-16.

6.47 Ideally, the CPI weights should be as up to date as possible and be broadly representative of the expenditure pattern that might be expected over the life of the index series. Thus, when the September quarter 2017 link was being introduced, it was necessary to consider whether any developments and policy changes between 2015-16 and the September quarter 2017 might have significantly affected the expenditure pattern and whether any revalued expenditures needed to be adjusted.

6.48 There were no major policy changes identified during this period that would have significantly changed volumes between 2015-16 and the September quarter 2017.

6.49 Items where expenditures were likely to have changed between 2015-16 and the September quarter 2017 were also investigated. This was the case with 'Audio, visual and computing equipment'. The 2015-16 HES data provided the latest information on household expenditure on these items which was price updated to the September quarter 2017. However for Audio, visual and computing equipment, this approach may lead to underestimation of the weight in the CPI due to the relatively high volume growth in the quality (size and features) of these high technology goods compared to other products in the CPI. Therefore, a volume increase of around 12% was calculated from the National Accounts HFCE components between 2015-16 and the September quarter 2017.

6.50 In most cases the adjustments to expenditure were made without compensating adjustments to other expenditure in the CPI basket. The implication is that changes in such expenditure were assumed to have come from or gone into savings.

## Lower level weights

6.51 Although the weights are expressed in terms of expenditure shares, it is not the expenditure shares (where expenditure is given by the product of quantity and price) that are held constant (or fixed) from period to period. What are held constant are the quantities of products underpinning these expenditures such as the number of litres of petrol purchased each period on average by households. Weights are presented in expenditure terms because it is not possible to present quantity weights in a meaningful way, e.g. the quantity of health services. The relative expenditure shares of items will change over time in response to changes in relative prices.

6.52 While the implicit quantity weights are held constant at the expenditure class level, the weights of items within an expenditure class (e.g. different types of bread) can be varied between periodic reviews to reflect changed purchasing patterns. Any weight changes are introduced into the CPI in such a way as to not affect the level of the index.

6.53 Information from reliable sources are used to assess the importance of one product relative to another. Sources include data collections from the ABS and other (both private and government) organisations, and other publications such as industry or market research reports. Information from the HES is also considered but, for the main part, is not sufficiently detailed or reliable at the lower levels of the CPI structure. For example, the HES data for types of appliances purchased would not be as reliable as industry sales data because of the small samples in the HES. The availability of transactions (scanner) data provides opportunities to use these data for weights at the lower level, particularly for the Food and non-alcoholic beverages group.

6.54 At the price sample or elementary aggregate level, there are no explicit weights. Rather, the price samples are constructed so they are self-weighting. For example, if there were a price sample for medium chocolate bars, and the major grocery outlets had 80% of these sales and convenience stores 20%, then the price sample would be selected so that for every price from a convenience store there are four prices from the major grocery outlets. Transactions data provides expenditure data at the elementary aggregate level which are used for weighting purposes. For more information see Use of transactions data in the Australian CPI of this manual.

## CPI weights using HFCE data for the 2018 update

6.55 HFCE measures expenditure by resident households on goods and services, whether the expenditure is made within the domestic territory or by Australian residents abroad, and expenditure by Non-Profit Institutions Serving Households (NPISH). Conceptually, the measurement of HFCE aligns closely with the HES. HFCE data captures household expenditure, including the Goods and Services Tax (GST). Both HFCE and HES data cover expenditure by Australian households only and exclude expenditure by non-residents in Australia.

6.56 The HES is used as a major benchmark in the compilation of the HFCE series. Benchmarking is a technique used by National Accounts, where less frequent (higher quality) data sources are used to validate more frequent estimates of HFCE. The other major benchmark used for HFCE is the Retail and Wholesale Industries (cat. no. 8622.0).

6.57 The use of HFCE data for CPI weights has many potential benefits for both internal and external users of inflation statistics, including more representative weights while at the same time aligning with international recommendations. The use of HFCE data will ensure the continued production of a robust CPI; and provide greater coherence between macro-economic statistics within the ABS.

6.58 For the 2018 update, movements in the HFCE data for the years 2015-2016 to 2016-2017 have been applied to the 17th series (2017) weights, which were primarily derived from the HES. The approach for the 2018 update to re-weight the CPI and SLCIs can be summarised as follows:

i. Align the HFCE data with the scope and classifications of the CPI and SLCIs at a detailed product level. This requires the removal of some components of HFCE (e.g. expenditure by Non Profit Institutions Serving Households).
ii. Produce a concordance of the HFCE data to the Consumer Price Index Commodity Classification (CPICC). This provides HFCE data for each CPI expenditure class (EC) for the CPI and SLCIs. These first two steps result in HFCE data that has been aligned to the same concepts and scope of the CPI and SLCIs.
iii. Calculate movements from 2015-16 to 2016-17 for the ‘HFCE aligned’ data from step (ii), and apply these movements to each CPI EC to update the expenditure values.
iv. Price update the 2016-17 expenditure values for each CPI EC to the September quarter 2018.
v. Re-scale the price updated expenditure values across all CPI ECs so they sum to 100.0 (i.e. expenditure shares).

For more information on the 2018 CPI weight update refer to Information Paper: Introduction of the Consumer Price Index Weight Update, 2018 (cat. no. 6470.0.55.002)