Mafia effect, higher taxes in infiltrated municipalities

A municipality without ever a shadow of a scandal runs undisturbed under the radar of controls. But the balance sheet doesn't lie: high expenditure on construction, sluggish collection and a rising tax burden. One anomaly after another that today an experimentally developed predictive algorithm is able to translate into an index of the risk that that apparently 'clean' municipality is actually infiltrated. This is not to say that the model accuses i Municipalities: its output, which will have to be refined, is a probability score, not a trigger for opening an investigation. And yet, the machine learning system of the Financial Intelligence Unit - the anti-money laundering body of Bankitalia headed by Enzo Serata - is proving reliable, as emerges from a study developed by Stefano Iezzi and Claudio Pauselli of the FIU.

Consider that it was accurate in analysing 6,771 municipalities, correctly distinguishing infiltrated and non-infiltrated in the period 2016-2021. Not only that. It has identified indices of alert in administrations that have not been touched by anti-mafia investigations in that timeframe. This is the case in Sicily, Calabria, Campania, and Apulia, but also in Lazio and Lombardy, where there are also municipalities with low levels of AML cooperation.

The machine learning system returned recurring patterns ('patterns') that emerged from the analysis of budgets. In Mafia-risk municipalities, for example, the tax system jams. Imu, waste tax (Tari), total revenue: collection collapses. It is a constant already reported by the most recent studies and finds confirmation in the numbers: where there is organised crime, municipal coffers empty. The causes? Two, mainly. On the one hand, the administration paralyses or turns a blind eye, avoiding collecting from friends and associates. On the other, citizens stop paying if they know that the money ends up in the wrong hands. Especially if the waste collection is managed by firms 'close' to the clans. But there is a paradox: the less taxes are collected, the higher the rates go up. In infiltrated municipalities, the per capita tax burden is higher. Who pays? The non-aligned: citizens and businesses out of the loop. In the meantime, the infiltrated administrations take advantage of financial autonomy to compensate for the poor collection, with extra-tax revenues: mulets, tariffs, sales. Easy to manage, difficult to track. This reduces dependence on state transfers and widens the room for manoeuvre to favour friends.

A recurring pattern of infiltrated municipalities concernsexpenditure. They are low for education culture, transport and welfare, while they go up for building and waste, sectors in which mafia organisations historically invest by laundering dirty money. The impact is also in procurement. Criminal organisations, we read in the algorithm's presentation documents, 'often engage in corrupt practices that inflate these expenditures, for example by awarding contracts to related companies at increased prices or through fraudulent schemes to divert public funds. The result is an increase in operating expenses through misallocation or waste of resources'. Infiltration, moreover, 'can cause an increase in personnel expenses through superfluous hirings, friends or affiliates. This creates rigidity as these expenses are fixed and difficult to reduce'.

In addition, 'corruption can lead to long-term contracts that are difficult to renegotiate or cancel, locking the administration into high costs; finally, corrupt administrators can contract unnecessary or inflated debts, often with opaque agreements or uncompetitive rates, increasing interest expenses'. In general, therefore, infiltrated municipalities 'suffer from aweakening of governance and control mechanisms, facilitating budget manipulations'. The indicator is in an experimental version that still needs to be worked on. But its future application could 'contribute to the development of more effective anti-mafia policies', the study notes.