There is no reason to suspect that the HIOA overestimates or underestimates expense items.
A savings rate of 20% under the HES would become a little under 8% if the HIOA adjustment is applied; a savings rate of 5% under the HES becomes negative 10%.
The extent of adjustment suggested here is probably excessive: it is unlikely that the adjustments calculated by comparing the aggregate HES and the HIOA data would apply exactly in any given case, and there remain some issues with HIOA as well.
Using data from the HIOA, he takes total tax payments and subtracts government transfer payments (such as national superannuation) and government consumption expenditure (such as health).
This has been in contrast to estimates from the HIOA, which, as we have seen, are now strongly negative.
Their paper has a comparison of the income from the HIOA and income from the HES.
However, a comparison of expenditure measures showed that HES expenditure was only 84 percent of HIOA expenditure.