Portal de Periódicos da CAPES

Extreme events and the resilience of decentralised governance

2023; Routledge; Linguagem: Inglês

10.1080/00343404.2023.2255627

1360-0591

María Cadaval Sampedro, Ana Herrero Alcalde, Santiago Lago Peñas, Jorge Martínez-Vázquez,

Fiscal Policy and Economic Growth

ABSTRACTExtreme events, such as economic crises, natural disasters or military conflicts, can affect the balance between centralisation and decentralisation forces across countries and transform, temporarily or more permanently, the design of multilevel governance. Using a panel of 91 developing and developed countries from 1960 to 2018, and another for Organisation for Economic Co-operation and Development (OECD) countries between 1995 and 2018, we examine the effects of extreme external shocks on the decentralisation level. We find that armed conflicts boost decentralisation, while natural disasters reduce it only in non-OECD countries, with long-lasting effects in both cases. Economic recessions do not have significant effects on the level of decentralisation, except for the lasting effects on expenditure recentralisation in OECD countries.KEYWORDS: fiscal decentralisationextreme eventsgovernanceresiliencepolitical decentralisationJEL: H60H71H77H84 DISCLOSURE STATEMENTNo potential conflict of interest was reported by the authors.Notes1. On the contrary, in the case of small-scale events, we could expect a larger role of subnational governments due to their proximity to the effects of the crisis, in line with Tselios (Citation2021) and De Mello and Tovar (Citation2022).2. Although we also considered the inclusion of total population as a regressor, the dynamics of the population are already captured by the dynamics of the population density (density is defined as the ratio between the population and a constant such as area), and the differences between countries in their population size (the average population in the sample period) are captured by the fixed individual effects.3. Controlling for income distribution with the Gini index was discarded because of the sharp decline in the sample size. The same applies for the use of measures of party system nationalization.4. The index based on expenditures at different levels measures the share of total public spending implemented by each level of government, where intergovernmental transfers are counted as financing part of subnational expenditures, but not as part of central government expenditures. As for the revenue index, it only includes for subnational governments their own tax collections and any other sources of income different from transfers.5. There is an inherent difficulty in measuring the costs of natural disasters over space and time, and this could obviously bias the dependent variable we are using. However, the EM-DAT database is widely accepted as a source when analysing the size and impact of those disasters.6. We checked the robustness of our results by separately estimating the effects of internal and external (international) conflicts, since the effects of the latter could be of the opposite sign. However, results turned out to be basically the same.7. Results for DENSITY could be explained by its low within-variation: most of its variation is across countries and not over time, The 'between standard deviation' of the variable is three times its 'within standard deviation'. Hence, 'the fixed effects will soak up most of the explanatory power of those slowly changing variables' (Beck, Citation2001).8. See Blackburne and Frank (Citation2007) for a more detailed technical discussion.Additional informationFundingThis paper benefitted from the funding of project PID2022-136276NB-C22, 'Decentralisation and Extreme Events in Contemporary Democracies' of the Ministerio de Ciencia e Innovación, 2023–26.