Wednesday, December 8, 2010

A few thoughts on National Accounts - Household Income

I stumbled upon a few things that I think are relevant to my first post on the price-income calculation debate that I wanted to share.  Perhaps the most relevant thing is that - despite commentators and the RBA saying that the national accounts measurement of household income is the 'most relevant' to be used for the purposes of calculating affordability of house prices - there seems to be considerable data quality concerns about this metric. 

For example, the ABS published the following in its 2007 item titled "Spotlight on National Accounts: Household Savings Ratio":

As the difference between the Household disposable income and Household final consumption expenditure is relatively small, caution should be exercised in interpreting the Household saving ratio in recent years, because major components of household income and expenditure may be subject to significant revisions. Many of these series are based on annual indicators and in a number of instances the household sector is measured as a residual of all other sectors. (emphasis added)

The most worrying revelation in the above quote from the ABS is that the methodology for calculating the income attributable to the 'household' sector in the National Accounts appears to be "a residual of all other sectors" in "a number of instances".  Income that doesn't fit neatly into 'government' or 'commercial' income is therefore simply attributed to households?  This seems totally bizarre to me.  This clearly also impacts the reliability and interpretation of the official national household savings rate (which is actually the context for the above ABS quote/paper).

When calculating affordability of prices for any asset or consumption item (including housing) this suggests (at least to me) a strong methodological reason to rely much more on the direct household income surveys conducted by the ABS which actually quantify household's income available for consumption, as opposed to National Accounts data.  As discussed in my earlier post these income levels can be easily adjusted for subsequent wage inflation if the 2-3 year lag in the data is of concern.