A1.1 Survey of Income and Housing (SIH)
The Survey of Income and Housing (SIH) is a national sample survey conducted by the Australian Bureau of Statistics at varying intervals from 1981–82 to the most recent survey in 2007–08.23 The sample is drawn from residents of private dwellings in both urban and rural areas of Australia.24 The SIH collects detailed information about the level and sources of income, and personal and household characteristics of people aged 15 years and above. In some cycles, information on other topics such as assets and liabilities, or household net worth, is also collected, as in 2007–08.25 The final sample size in 2007-08 was 9,345 households and the 18,326 people living in these households. Data are collected during face to face interviews by trained ABS interviewers using computer assisted questionnaires at both household and individual levels (ABS 2009).
The SIH series provides good historical data on household incomes and housing costs and is a useful source of national data on home ownership over time. As with any sample survey, there are limitations. The sample size limits the potential for analysis on a spatial basis except for a capital city and ‘rest of state’ comparison although in this study, because of the need to keep the sample size large enough for meaningful analysis of other variables, no analysis below the national level was undertaken.
SIH data on housing costs refer to mortgage repayments and property rates only; the SIH does not include data on other ongoing costs of home ownership, for which we have to use the HES. Mortgage repayments are included in housing costs for owner occupied dwellings if ‘the purpose of the loan when it was originally taken out was primary to build, buy add to or alter the occupied dwelling’ (ABS 2006: 27). As with other ABS sample surveys, data items and data definitions are subject to change over time.
SIH data are available in Confidentialised Unit Record Files (CURFs) consisting of unidentified individual statistical records containing data on persons belonging to income units in private dwellings, including state and capital city/rest of state identifiers. Data as they relate to this study include household, family and income unit types; age, details of mortgages and loans, housing costs, type of tenure and landlord, type of dwelling structure, and details of weekly and annual income by source of income for persons and income units.
Data in the SIH like other ABS sample surveys are weighted to enable inference to the general population from which the sample is drawn. To do this, ‘a “weight” is allocated to each sample unit e.g. a person or a household. The weight is a value which indicates how many population units are represented by the sample unit’ (ABS 2009: 34). Weights for the 2007–08 SIH were based on the ABS Census of Population and Housing 2006. The combination of the sampling method and
23
The SIH was previously the Survey of Income and Housing Costs (from 1994-95) and prior to that was the Survey of Income and Housing Costs and Amenities (from 1981-82). The survey was conducted at varying intervals until 2003-04 since when it has been conducted biennially.
24
The sample covers 97 per cent of the Australian population; it excludes very remote areas (ABS 2009: 26).
25
Details about the SIH 2007-08 (Information Paper, Questionnaire and Prompts) can be found at http://www.abs.gov.au/AUSSTATS/[email protected]/DetailsPage/6553.02007-08?OpenDocument.
weighting process means that we can generalise from the SIH sample to the general population with some confidence.
A1.2 Household Expenditure Survey (HES)
The Household Expenditure Survey (HES) is a national sample survey conducted by the Australian Bureau of Statistics at intervals from 1975–76.26 The HES collects information by interviewing usual residents of selected private dwellings in urban and rural areas of Australia, excluding very remote areas, and those selected also fill out a personal expenditure diary over a two week period. A different group of households is selected for each survey; in other words the HES is not a household panel survey. The 2003–04 HES was conducted jointly with the SIH, with the HES sample being a sub-sample.27 It contains very detailed information about household expenditure across many items and is used for various purposes, including weighting of types of expenditures for calculation of the Consumer Price Index (CPI) as expenditure patterns change over time. The HES collects information with respect to households and all the people comprising those households. It is therefore possible to produce aggregate data from the surveys to an individual person, income unit or household level (ABS 2006: 24).
Sample numbers vary somewhat over successive HES, with the 2003–04 final sample comprising 6,957 households. It is not possible to do spatial analysis (other than metropolitan/non-metropolitan) using the HES due to the sample size. Because of the need to keep the sample size large enough for meaningful analysis of other variables no analysis below the national level was undertaken in this study.
As with other sample surveys, data items and definitions are subject to change as expenditure patterns change.
HES data are available in CURFs consisting of unidentified individual statistical records containing data on persons belonging to income units in households occupying private dwellings, including state and capital city/rest of state identifiers.
A1.3 Victorian Valuer-General’s Property Sales Statistics
The Victorian Valuer-General’s Guide to Property Values is published annually. It provides sales data for houses, flats and units, vacant residential land, and commercial, industrial and rural properties. The guide provides property sales statistics for Victoria’s 79 municipalities and lists yearly medians by suburb covering houses, units and vacant blocks. It provides statistics on municipality trends as well as state-wide totals. The Guide is derived from detailed data that provide prices for all land/property transactions recorded for stamp duty purposes in the state in any one year, categorised into detached dwellings and units/apartments and blocks of land. The data are obtained from Notices of Acquisition which are required to be completed by each purchaser of a property within one month of the acquisition of any real estate in Victoria and on which stamp duty is payable. The information is collated and loaded to a master property file. The data are available in hard copy format providing details of property values for local government areas and suburbs. They are also available in CURF electronic format which provide more detailed data but from which the names of purchasers are stripped from the data file. Individual addresses are retained. Almost all the analysis in this study involved analysis of the CURF property sales data rather than the hardcopy Guide to Property Values.
26
The HES was conduced in 1975-76, 1984-85, 1988-89, 1993-94, 1998-99 and 2003-04. 27
Despite the requirement for notification within one month of acquisition, there are often delays in recording data, and the most recent data can be amended some months later. For that reason we have not used any 2009–10 data. Some data are also subject to a time lag due to the operation of the stamp duty system, for example, an apartment sold in 2006 off the plan may not be recorded until stamp duty is paid in 2008. This means that the data are constantly being adjusted. Despite this difficulty which mainly occurs in respect of larger blocks of units/apartments, Valuer-General’s data are the most accurate source of housing prices as they are based on a ‘population’ rather than a sample survey which is subject to sampling error.
Residential sales since the 1990s have been allocated to one of 13 categories embracing houses, a range of multi-unit forms (flats/units, block flats, town house, retirement dual occupancy) and other property types including vacant land. In earlier years, such as 1981 the year used in this study for comparative purpose, there were only two dwelling categories: houses and ‘own your own’ flats. For this reason, unless specified, we use data for houses as this has been a consistent series. However, over time, the percentage of sales of dwellings other than houses has increased as the dwelling stock has been diversified. In 2008 there were 77,987 dwelling sales recorded in Victoria of which 66 per cent were houses.
A1.4 Census of Population and Housing
The census is the largest statistical collection undertaken by the Australian Bureau of Statistics. The first was held in 1911 and since 1961 a census has been taken every five years, with 2006 being the latest. The scope of the census is all people in Australia on census night, excluding foreign diplomats and their families. The smallest spatial unit for collection and analysis is the census collectors’ district of which there were more than 38,000 in 2006. These can be regrouped into different types of geographical area for research purposes. In this study 2006 census data were used mainly for mapping purposes as it is the only data source on household income that enables metropolitan wide spatial analysis.