Silent fiscal decentralization: the increase in spending by states and municipalities after the pandemic

The Brazilian fiscal debate has revolved almost entirely around the federal government: primary targets, fiscal framework, Bolsa Família, IRPF exemption, gross debt. A relevant agenda, but one that hides one of the main structural changes in public spending after the pandemic: the sharp increase in state and municipal expenses, a phenomenon that was called "silent fiscal decentralization" by Manoel Pires, my colleague at FGV Ibre.

From the perspective of national accounts (that is, GDP), in which each expenditure is attributed to the sphere that actually executes it, the primary expenditure of the "budgetary" central government was accommodated around R$100 billion above the extrapolation of the 2015-19 trend — a deviation that roughly corresponds to the expansion of Bolsa Família from 2021 onwards. Spending by regional governments jumped from around R$2 trillion to R$2.7 trillion, in real terms, since 2022. Of the increase in general government spending (federal government plus states and municipalities) since 2019, around 70% occurred at the regional level.

It is worth remembering that a non-negligible part of what the National Treasury reports as federal expenditure is, in fact, transfer to subnational entities: supplementation to Fundeb, transfers from SUS, FCDF, Kandir Law, part of parliamentary amendments, among others. The total transferred by the Union to regional governments, considering revenue sharing and transfers via primary expenditure, went from 6.2% of GDP on average from 2006-19 to almost 8% of GDP since 2022.

This fiscal decentralization brings three sets of problems.

The first is coordination with monetary policy. The Central Bank reacts to the aggregate fiscal impulse, not just that of the Union: between 2022 and 2025, the regional spending impulse totaled around R$700 billion, almost 70% above the central government's R$400 billion.

The second is sustainability. Only the Union is charged with generating primary compatible with the stabilization of public debt. Regional governments owe above all to the Treasury itself, which renegotiated these debts in 1997, 2014, 2016 and, now, with Propag (2025). There is no market discipline, and the institutional brakes are limited to LRF personnel spending ceilings, which can be circumvented by creative accounting, and debt limits (which are conditioned by Capag's rating, which is highly pro-cyclical). The profits from the bonanza became permanent expenditure, without savings; losses, however, are socialized.

The third is the effect on the Union's own ability to meet its primary outcome targets. In terms of expenditure, transfers to regional governments have grown significantly since 2020. In terms of revenue, half of the IR and IPI collection is automatically distributed via FPE, FPM and regional funds — it is worth noting that the FPM percentages have been frequently raised by Congress, as was the case in 2021.

The Fiscal Responsibility Law (LRF) of 2000 was a landmark, but it was designed for another country and failed in several aspects. We need to improve it, especially with regard to greater accountability for fiscal balance, which also needs to reach regional governments and other Powers.

Congress expanded Fundeb and FPM and converted the amendments into a decentralized spending policy of R$50 billion per year, with no counterpart in fiscal results. Legislative and Judiciary, in all spheres, continue to decide readjustments and compromises as if the budgetary restriction were an exclusive problem of the Executive.

The fiscal consolidation that Brazil needs to implement cannot come only from the federal Executive.