John Garnaut has a piece in today’s SMH on tax evasion. I’m quoted on one of my pet peeves – that Australia has never conducted a randomised audit of all taxpayers. We’ve done random audits of particular industries – such as the building sector – but never of all taxpayers. Why? I think it’s because our politicians are scared of what they might find, but I’m open to hearing alternative explanations.
As a result of never having done a random audit study, we don’t know something as simple as whether tax evasion (as a fraction of income) is higher for the rich or the poor, the young or the old, urban taxpayers or rural taxpayers. And the answers could go either way. Garnaut cites findings from a NBER working paper by Joel Slemrod and Shlomo Yitzhaki on the US random audit program in the 1980s, which found that on a proportionate basis, tax evasion was higher at the bottom end.
The 1988 audit – the last one before the program was banned by Congress – found taxpayers earning over $US100,000 ($136,000) reported 96.6 per cent of their true income, while those earning less than $US25,000 reported 85.9 per cent.
Wanted: an Australian politician who understands randomised trials. Apply here.