Seven tax return labels that quietly attract ATO attention

The ATO's data matching gets better every year. Most queries are not triggered by aggressive positions; they are triggered by simple mismatches between what is on the return and what the Tax Office already knows. These are the labels where a quick second look pays off.

The seven to watch

• Work related expenses. Round numbers, amounts sitting right at a threshold, and claims that jump from last year all draw attention. The question is always the same: can it be supported?
• Rental property deductions. Repairs versus capital improvements, interest apportionment, and the periods a property was genuinely available for rent are perennial trouble spots.
• Motor vehicle claims. Logbook versus cents per kilometre, and claims that sit suspiciously close to the cap, are easy to flag.
• Capital gains. Missing or mis-timed CGT events, forgotten cost base elements, and disposals that appear in third party data but not on the return.
• Dividend and interest income. Pre-fill makes omissions obvious. A figure that does not match the reported data is a straightforward mismatch.
• Private health insurance and offsets. Small details, but the ATO holds the matching data, so an error here is quickly visible.
• Trust and partnership distributions. Amounts that do not reconcile across the linked entities are a common and avoidable flag.

The common thread

Notice what these have in common. None of them require judgment about an aggressive position. They are consistency checks: does the label tie back to the workings, and does it match the data the ATO already holds?

A simple habit

Before lodging, it is worth asking three questions of each significant label. Does it reconcile with the working papers? Is it consistent with prior years, and if not, is there a reason? Would it match the data the ATO is likely to hold? A return that passes those three tests rarely produces a surprise.

This is general information, not advice about any particular return. As a pre-lodgement habit, though, a quick pass over these seven labels catches most of the errors that would otherwise surface after the fact.