Unlike other products on the market, CSP was designed using a top-down method. Precisely started with the overall market, then apportioned the total of each SA1 by category. The benefits of this approach include:
- SA1 estimates roll up to an established benchmark. Precisely first determines the size of the market by state or territory, before any modeling begins. This ensures that estimates are part of a consistent and robust benchmark. We recognize that there are inherent problems with the overall amount of spending when rolling up averages from the Household Expenditure Survey. Therefore, we put considerable effort into calculating market size for each of the categories in CSP.
- This method overcomes several biases in the data for household expenditure. The Household Expenditure Survey from ABS is designed to provide weights for the consumer price index (CPI). As such, the scope of households surveyed is not proportional to the spending profile of Australian households. Lone-person households make up 30% of households surveyed, but only 13% of spending. Couples with children make up 37% of spending, but only 19% of households surveyed. Households headed by persons 65 and older make up 33% of households surveyed, but only 12% of total spending. Finally, the survey relies too heavily on smaller metropolitan areas. The Sydney and Melbourne metropolitan regions are somewhat under-surveyed relative to their share of spending, while Adelaide, Perth, Hobart, Darwin, and Canberra are all over-surveyed.
- New versions of CSP are comparable with order versions. Starting with the 2020 release, the level of detail in expenditure sub-categories increased; however, the higher categories are still comparable. Because Precisely begins with market size, new versions of Consumer Spend Potential evolve from one release to the next.
Market size for each CSP category is calculated from a variety of sources, the primary being Household Final Consumption Expenditure (HFCE) tables from the National Accounts (Australia's gross domestic product figures). Because this dataset is an integral part of GDP statistics, it is the final arbiter of household market size in Australia. The only downside to using GDP figures in CSP is that they are presented at a relatively high level. For example, the Expenditure tables only report on total food expenditure and do not break it down into anything more granular.
For this reason, Precisely uses other sources to augment HFCE data. These include other National Accounts tables from the Australian Bureau of Statistics as well as Retail Trade, Apparent Consumption of Alcohol, International Trade in Services, and Socio-Economic Indexes for Areas (SEIFA). We also incorporate data from other Commonwealth sources – the Australian Prudential Regulation Authority, the Productivity Commission, and Tourism Research Australia – as well as other publicly available sources.
Various sources are also used when conducting quality checks. Online Spending is an example that has seen significant changes since the COVID-19 pandemic. Numerous articles and publications are used at a high level to test modeling and provide indications of where any additional input sources may be required.
The sum total of these data sources provide information about:
- Household income
- Number and percentage of persons aged 5 to 14
- Number and percentage of persons aged 20 to 34
- Number and percentage of persons aged 65 and over
- Number and percentage of households paying off a mortgage
- Number and percentage of households paying rent
- Number and percentage of households who have moved within the last five years
- Number and percentage of persons who have gone to, or graduated from, university
This top-down approach produces estimates that are robust and transparent, and provides a stable set of figures on which businesses and organizations can rely now and in the future.