Decentralization and intra-country transfers in the Great Recession: the case of the European Union

ABSTRACT

Empirical evidence is found supporting the assignment of the redistribution function to the central government. Viewing the Great Recession, or recessions in general, as a shock, it is found that after the shock, more decentralized economies show a more negative response of subnational government social protection expenditures than more centralized ones. Results for the vertical fiscal imbalance are consistent with the theory of soft budget constraints: economies with a greater vertical fiscal imbalance show larger increases in subnational government social protection expenditures after the shock and greater prior subnational government borrowing. Caution should be used in extending the argument to the European Union level.

KEYWORDS:

• decentralization
• transfers
• social protection
• federalism
• redistribution

JEL:

• H12
• H77

ACKNOWLEDGEMENTS

The author thanks Carlos Cuerpo; the participants at the workshop on ‘Decentralization after the Great Recession: Fine-Tuning or Paradigm Change?’, held in Santiago de Compostela, Spain, October 2017; the participants and the author’s discussant, Antonio Afonso, at the 2018 International Institute of Public Finance (IIPF) Meetings held in Tampere, Finland, August 2018; the participants and the author’s discussant, Louis-Philippe Beland, at the 111th Annual National Tax Association Meetings held in New Orleans, Louisiana, United States, November 2018; four anonymous referees; and Jorge Martínez-Vázquez for helpful and stimulating comments and discussions. Yaoqi Lin provided timely and valuable research assistance.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the author.

Notes

1. Lago-Peñas, Martínez-Vázquez, and Sacchi (2019) find that budget rules may be an important control for stability. The present fixed country effect regressions control for such rules that did not change over the sample period (or with year-fixed effects that changed for all countries in a given year). To the extent that regulatory changes are related to the business cycle, inclusion of the unemployment rate is also a useful control.

2. Beramendi and Rogers (2019) address a somewhat different but related issue. Some of their results are consistent, namely that more fiscally decentralized nations have more inequality and less redistribution.

3. The present paper concentrates on redistribution in this literature review. Carlino and Inman (2013) find evidence supporting the positive spillover argument for stabilization in the United States. See also the aforementioned paper by Lago-Peñas et al. (2019), the literature cited by Feld, Schaltegger, and Studerus (2018) as well as their analysis for Switzerland, and Buettner (2002) for evidence for Germany.

4. See also Schmidheiny (2006), who extends this type of model to a progressive income tax with heterogeneous preferences over public goods.

5. See also Eugster and Parchet (2019) for evidence on tax competition in Switzerland using a novel identification method that takes advantage of language differences.

6. The literature on expenditure competition started perhaps with Case, Rosen, and Hines (1993). For recent surveys on tax competition see, for instance, Keen and Konrad (2013) and Devereux and Loretz (2013).

7. See https://ec.europa.eu/eurostat/data/database. The figures for social protection spending are not by level of government, and thus are for all levels. SNG social protection spending is calculated here by subtracting expenditures on health, unemployment and old-age pensions, as explained in the text.

8. See http://www.oecd.org/tax/federalism/fiscal-decentralisation-database.htm#C_4. The data are consolidated.

9. See http://www.cpds-data.org/.

10. Regressions were rerun by defining recession as one quarter of negative growth. The results are broadly similar, so these results are omitted.

11. However, recall the data limitation that this is for the national parliament, which could explain its insignificance.

 

Additional information

Funding

This work was supported by PSC-CUNY [grant number 61605-0049].