Who absorbs the shock? An analysis of the fiscal impact of the COVID-19 crisis on different levels of government

Abstract

This paper discusses the fiscal impact of the COVID-19 crisis across levels of government. It contrasts the composition of revenues and expenditures of different levels of government and their main institutional differences. For revenues, an error correction model is used to estimate the short-term elasticities of different levels of government, showing that subnational governments’ (SNGs) revenues are less sensitive to economic downturns than central governments’, mostly because SNGs tend to rely more on recurrent taxes on immovable property. For expenditures, central governments are often responsible for the bulk of expenditure on social protection, while SNGs have a higher investment-to-revenue ratio. The combination of these differences of expenditure assignment with the substantial budget and borrowing constraints that SNGs face creates a tendency towards pro-cyclicality at the subnational level and counter-cyclicality at the central level. Furthermore, in the context of the COVID-19 crisis, in order for SNGs to have the fiscal capacity for necessary measures to tackle the outbreak, central governments supported SNGs mostly through intergovernmental grants. As a consequence, central governments have absorbed most of the shock. Supplementary Information The online version contains supplementary material available at 10.1007/s10368-021-00518-1.

The intergovernmental fiscal outlook and the implications of Russia’s war against Ukraine, high energy prices and inflation

“…First, SNGs tend to rely more on recurrent taxes on immovable property, which has a relatively smaller tax buoyancy (i.e., sensitivity to economic growth) because its tax base is, by and large, property values (Dougherty & De Biase, 2021). Second, central governments bear more responsibilities than SNGs in stabilising economies and, thus, tend to spend more on fiscal support aimed at alleviating the impact of crises (for a detailed discussion, see De Biase .…”

“…There are, though, substantial asymmetries across countries -SNGs buoyancy coefficients tend to vary from 0.68 to 1.88, 9 meaning that the impact of a reduction in economic growth can be up to nearly three times higher for SNGs in one country compared to others. These variations are mostly driven by differences in their tax mixes, with the impact being more substantial for SNGs relying on corporate income tax revenues and less substantial for those who rely mostly on property taxes 10 (for more details, see Dougherty & De Biase, 2021).…”

“…A buoyancy of 1.1 means that when GDP increases by 1% revenues are going to increase by 1.1%. For a detailed discussion of SNGs long-term buoyancies see (Dougherty, De Biase, et al, 2022) and for an analysis of SNGs short-term buoyancies see (Dougherty & De Biase, 2021). 9 Values estimated considering the mean group estimator for each type of tax (Dougherty & De Biase, 2021).…”

“…Another additional risk comes from the fact that central governments are in a significantly worse position than SNGs at the moment. Not only have they absorbed most of the COVID-19 shock (Dougherty & De Biase, 2021), but they are also absorbing the fiscal costs of cushioning households' living standards at a time of high energy and food inflation and, in some countries, are planning to increase defence spending.…”

Improving subnational governments’ resilience in the wake of the COVID-19 pandemic

“…A recent empirical analysis of OECD and selected partner countries' tax buoyancies (i.e., changes in revenue associated with fluctuations in the business cycle) suggests that, on average, subnational revenues tend to indeed be less cyclical than national revenues (Figure 3) (Dougherty and de Biase, 2021). The cyclicality of subnational revenues is also positively correlated with the degree of dependence of SNGs on intergovernmental transfers.…”