Regional tax commission, judgment 83/2012, confirming the first-instance decision 17/01/11 of the provincial commission of Lodi. The Revenue's appeal was dismissed.
What the case was about
The Revenue assessed a professional's income for 2005 because the income declared diverged from the figure produced by the studi di settore — statistical models that estimate what a business of a given kind, size and location ought to earn. A demand for payment followed.
The taxpayer challenged both, and won at first instance. The Agency appealed.
What the court decided
The appeal was dismissed. Proving the existence of serious inconsistencies between declared income and the statistical estimate is for the Revenue. The divergence between a declaration and a model is a starting point for enquiry, not a finding of fact that the taxpayer must then disprove.
A related point had already been settled below: the statistical parameters do not apply to a professional who has not succeeded in recovering fees through the courts. Income that has been billed but never collected is not income.
Why it matters
This is the limit of the presumption, and it is the single most useful thing to know about Italian tax litigation. The administration may begin from a statistical model; it may not end there. Somewhere it must show the real facts of the real business, and the point at which it fails to do so is where the assessment falls.
The models themselves remain opaque — the algorithm that produces the expected figure has never been fully disclosed — which is precisely why the burden of proof matters so much.