"Again Mr Joye quotes the spurious dwelling price-to-income ratio adopted by Rismark, the RBA and the banks that: (1) wrongly compares median dwelling prices against average household disposable incomes (it should use median incomes, which are significantly lower than the average); and (2) includes a number of items as income that are not actually available to be used to fund current consumption, such as superannuation contributions and imputed rents."
The problem with this is that Rismark compares average dwelling prices to average incomes (not just medians to averages), and I am pretty sure Leigh knows this. In Rismark's quarterly dwelling price-to-income ratio, we use the only regularly published disposable household income measure made available by the ABS, which is the quarterly National Accounts estimate of AVERAGE household disposable incomes.
To compare like-with-like, we calculate AVERAGE national dwelling prices across all regions and property types (see chart). We also provide a median dwelling price measure, which is not our preferred benchmark, for comparative purposes. We then produce a price-to-income ratio using averages on averages, which is how we get our 4.5x estimate.
Finally, we provide an estimate of how this ratio changes when one strips out owner-occupied rents from the ABS measure of disposable household incomes, which is the main non-cash item (the impact is trivial, as the chart below shows). I have published these charts here many times before, but since Leith seems to ignore them, I have posted them here again for his benefit.
