Skip to main content
Log in

Environmental Disasters and Electoral Cycle: An Empirical Analysis on Floods and Landslides in Italy

  • Published:
Environmental and Resource Economics Aims and scope Submit manuscript

Abstract

The aim of this paper is to analyse potential drivers of land use policy, in the form of building permits issued in Italian provinces. We first derive testable implications on the basis of a standard political agency framework, augmented to account for the impact of past environmental disasters (floods, landslides and earthquakes) and for the relevance of “building permits intensive” sectors in determining voters’ support to an incumbent politician. We then perform an empirical analysis that tests theoretical predictions using a unique dataset covering Italy in the period 2001–2012. Our main conclusions show that the occurrence of floods and earthquakes decreases building permits, implying that a bad history in terms of these phenomena strengthens the importance of voters affected by past disasters. No corresponding evidence seems to emerge with reference to landslides. On the other hand, the relevance of the construction sector increases the number of building permits issued. Finally, when elections approach, the number of building permits issued grows, suggesting that incumbent politicians may distort land use policies in order to favour “brown” voters in periods close to elections.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6

Similar content being viewed by others

Notes

  1. http://land.copernicus.eu/pan-european/corine-land-cover.

  2. See ISPRA (2015) and ISPRA (2017); data from the CORINE Land Cover dataset (European Environment Agency) are slightly different but coherent.

  3. http://esdac.jrc.ec.europa.eu/themes/landslides.

  4. Obviously, the shift from simple exposure to the actual occurrence of a catastrophic event is related to some natural and/or climatic features (Peruccacci et al. 2017). However, this does not enter our study, which deals with land use, exposure perceived by voters, and the ways in which these are expected to affect the chances that the incumbent local politician will be re-elected.

  5. Thus, in the terminology of Besley (2006), we investigate political failure leading to the misallocation of natural resources.

  6. In our setting, ideological green voters (e.g. environmentalists) are included in the ideological voters group (fraction \(1-\eta\) of voters). In the empirical part, we also control for the share of ideological green voters. We wish to thank a referee for pointing out this issue in our definition of “green voters”.

  7. We implicitly assume that both politicians and citizens are aware of the causal links among soil sealing, land instability, and extreme events, such as floods and landslides. Thus, we assume that all agents know about both risks and benefits of land exploitation. Obviously, we do not claim that politicians and citizens are expert geologists. Even so, the issue of hydro-geological risk is a matter of debate in newspapers and the media, and we can expect both the general public and policy makers to possess at least some basic knowledge.

  8. As clarified in our empirical analysis below, in Italy, local politicians set soil protection expenditure according to an exogenous national budget, which supports the assumption that r is exogenous in our model.

  9. Notice that, like in List and Sturm (2006), we assume that majority voting takes place.

  10. To simplify the exposition, and without loss of generality, we normalize \(\lambda =1\).

  11. We assume second order conditions for an interior maximum to hold, namely \(\left( \frac{f \sigma \eta }{(1-\eta )}\theta _{pp}-\frac{\partial ^{2}c}{\partial p^{2}}\right) <0.\)

  12. We consider only 96 of the current 107 NUTS3 provinces because of the creation of new provinces and related changes in boundaries in 2005 (affecting the provinces of Olbia-Tempio, Ogliastra, Medio Campidano, Carbonia-Iglesias, Cagliari, Nuoro, and Sassari) and 2009 (affecting the provinces of Monza and Brianza, Milano, Fermo, Ascoli Piceno, Bari, Foggia, and Barletta-Andria-Trani).

  13. On average, an Italian province has about 560 thousand residents.

  14. Unfortunately, there is no information on demolitions of buildings in Italy, with the sole exception of the 2016 Annual Report by FederCostruzioni, which mentions “...improvements in the building quality and energy performance through demolition and reconstruction of buildings ...coherently with what is happening in Europe” (FederCostruzioni 2016, p. 89, our translation). We interpret this as evidence, albeit tentative, that demolishing and reconstructing buildings is still not common practice in Italy. Thus, we cannot exclude that at least a portion of permits may be issued for buildings that replace existing buildings, rather than for new land development. This issue introduces a certain measurement error and may reduce the precision of our results. As second best, we exclude permits for extensions to existing buildings and consider only permits referring to new buildings (either on new land or on previously developed land). On average, this variable accounts for about 82% of total building permits. We believe new buildings to be more strongly correlated with actual land development and, to test this assumption, we use data from the CORINE Land Cover database (source: European Environment Agency—https://land.copernicus.eu/), available for the EU for years 2000, 2006, and 2012. For each province, we calculate the share occupied by man-made buildings in each of the 3 years and its increase in the two available time windows, 2000–2006 and 2006–2012. We then correlate this change in land use with the cumulated amount of building permits over the same time windows (in square meters, rescaled for the total area of the province). The estimated correlation coefficients are all positive and strongly significant. More specifically, the correlation coefficients between building permits for new buildings and changes in soil sealing are + 0.5166 (2000–2006) and + 0.5169 (2006–2012), suggesting that our dependent variable is a good proxy for increases in soil sealing.

  15. The intensity of past disasters is rather heterogeneous, with a large number of relatively small events and few major events. We consider only major events, since they are more likely to influence non-ideological green voters. In the absence of an objective measure of magnitude for what concerns floods and landslides, we follow the general approach found in commonly used databases, such as the EM-DAT repository, and adopt the presence of at least one fatality caused by the event as our selection criterion. The idea is that, through the local and national media, citizens are better informed about disasters that cause fatalities than about events with no fatalities. On the contrary, for what concerns earthquakes, we have detailed information on the actual magnitude of each event and set our threshold at 5ML (Richter scale).

  16. The variable is computed as: \({\textit{Stock disast}}_{i,t}=\sum _{s=1995}^{t}Disaster_{i,s}(1-\delta )^{s}\) where \(\delta =0.2\).

  17. As argued in the introduction, earthquakes differ from floods and landslides as there is no link between human activities and hazard but only between these and exposure (we take into account the intensity, not the damages or fatalities). We include earthquakes to control for their role in causing landslides (Peruccacci et al. 2017). On the contrary, when considering floods and landslides, human activities (e.g. soil sealing)—and the actions of politicians in particular—directly influence the hazard as well as the level of exposure to the hazard.

  18. This indicator is available for the period 2004–2012 and is built by combining different dimensions of the rule of law: crimes against the person or property, magistrates productivity, length of trials, tax evasion, and shadow economy. In our econometric analysis, we take the value of the indicator for 2004 and interact it with a linear time trend to allow provinces with different levels of rule of law to have different trends in building permits.

  19. Another possible source of endogeneity lies in the fact that our dependent variable also considers permits for buildings that help reduce the probability of fatalities when disasters occur (e.g. hospitals). However, these concerns are mitigated by controlling for province fixed effects and by introducing disasters with a lag.

  20. High hazard means that the return period of a flood is between 20 and 50 years; medium hazard refers to a return period of 100–200 years; lastly, low hazard areas have little probability of being flooded.

  21. In this figure we just consider the stock of floods and landslides as the hazard related to earthquakes is considered as exogenous.

  22. The F-test on the excluded instruments in the first stage is above the usual threshold of 10, suggesting that the instrument is strong.

  23. To saturate the model, we also include in the regressions interaction terms with these proxies and year dummies to account for potentially different trends in provinces which systematically differ in these dimensions.

References

  • Besley T (2006) Principled agents? The political economy of good government. Oxford University Press, Oxford

    Google Scholar 

  • Besley T, Burgess R (2002) The political economy of government responsiveness: theory and evidence from India. Q J Econ 117(4):1415–1451

    Article  Google Scholar 

  • Burgess R, Hansen M, Olken B, Potapov P, Sieber S (2012) The political economy of deforestation in the tropics. Q J Econ 127(4):1707–1754

    Article  Google Scholar 

  • Cutter SL, Barnes L, Berry M, Burton C, Evans E, Tate E, Webb J (2008) A place-based model for understanding community resilience to natural disasters. Glob Environ Change 18:598–606

    Article  Google Scholar 

  • Deng G, Gan L, Hernandez MA (2015) Do natural disasters cause an excessive fear of heights? Evidence from Wenchuan earthquake. J Urb Econ 90:79–89

    Article  Google Scholar 

  • Drury AC, Olson RS (1998) Disasters and political unrest: an empirical investigation. J Conting Crisis Manag 6(3):153–161

    Article  Google Scholar 

  • Esteban J, Ray D (2006) Inequality, lobbying, and resource allocation. Am Econ Rev 96(1):257–279

    Article  Google Scholar 

  • FederCostruzioni (2016) Rapporto 2016 - Il sistema delle costruzioni in Italia. http://www.federcostruzioniweb.it/index.php?option=com_content&view=article&id=1204:rapporto-2016-federcostruzioni&catid=31:archivio-news&Itemid=281

  • Ferraz C (2007) Electoral politics and bureaucratic discretion: evidence from environmental licenses and local elections in Brazil. Unpublished Manuscript, The Institute for Applied Economics Research, Brasilia, Brazil

  • Feyen L, Rutger D, Bodis K, Salamon P, Barredo JI (2012) Fluvial flood risk in Europe in present and future climates. Clim Change 112:47–62

    Article  Google Scholar 

  • Grossman GM, Helpman E (1994) Protection for sale. Am Econ Rev 84(4):833–850

    Google Scholar 

  • Guzzetti F, Reichenbach P, Cardinali M, Galli M, Ardizzone F (2005) Probabilistic landslide hazard assessment at the basin scale. Geomorphology 72:272–299

    Article  Google Scholar 

  • Haase D, Nuissl H (2007) Does urban sprawl drive change in the water balance and policy? The case of Leipzig (Germany) 1870–2003. Landsc Urb Plan 80:1–13

    Article  Google Scholar 

  • Hallegatte S (2012) An exploration of the link between development, economic growth, and natural risk (English). Policy Research working paper no. WPS 6216. World Bank, Washington, DC

  • Hallegatte S (2014) Natural disasters and climate change. An economic perspective. World Bank, Washington, DC

    Google Scholar 

  • Healy A, Malhotra N (2009) Myopic voters and natural disaster policy. Am Polit Sci Rev 103:387–406

    Article  Google Scholar 

  • Hervas J (ed) (2003) NEDIES Project - Lessons learnt from landslide disasters in Europe, Report EUR 20558 EN  

  • ISPRA (2015) Il consumo del suolo in Italia. Report 218, Italian Institute for Research and Environmental Protection (ISPRA)

  • ISPRA (2017) Il consumo di suolo, dinamiche territoriali e servizi ecosistemici. Report 266, Italian Institute for Research and Environmental Protection (ISPRA)

  • List JA, Sturm DM (2006) How elections matter: theory and evidence from environmental policy. Q J Econ 121(4):1249–1281

    Google Scholar 

  • Mysiak J, Testella F, Bonaiuto M, Carrus G, De Dominicis S, Ganucci Cancellieri U, Firus K, Grifoni P (2013) Flood risk management in Italy: challenges and opportunities for the implementation of the EU Floods Directive (2007/60/EC). Nat Hazards Earth Syst Sci 13:2883–2890

    Article  Google Scholar 

  • Nifo A, Vecchione G (2014) Do institutions play a role in skilled migration? The case of Italy. Reg Stud 48(10):1628–1649

    Article  Google Scholar 

  • Pailler S (2018) Re-election incentives and deforestation cycles in the Brazilian Amazon. J Environ Econ Manag 88:345–365

    Article  Google Scholar 

  • Peruccacci S, Brunetti MT, Gariano SL, Melillo M, Rossi M, Guzzetti F (2017) Rainfall thresholds for possible landslide occurrence in Italy. Geomorphology 290:39–57

    Article  Google Scholar 

  • Rogoff K (1990) Equilibrium political budget cycles. Am Econ Rev 80(1):21–36

    Google Scholar 

Download references

Acknowledgements

We are grateful to IRPI (Research Institute for Geo-Hydrological Protection) for data and support, and to Thomas Bassetti, Claudio Petucco, Massimiliano Mazzanti, and audience at the 2016 IAERE and EAERE Conferences for very useful comments and suggestions. We acknowledge financial support from the research project “La valutazione economica dei disastri economici in Italia” funded by the Fondazione Assicurazioni Generali. The usual disclaimer applies.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Alessio D’Amato.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendix: Additional Robustness Checks

Appendix: Additional Robustness Checks

As an additional robustness check, we account for the fact that for some specific area the NUTS3 level of analysis may be too aggregated because of the presence within the same province of locations with very heterogeneous geomorphological characteristics. For each province, we calculate the coefficient of variation (ratio between standard deviation and mean) of the within-province between-municipality measures of risk exposure (low, medium and high, respectively). The idea is that provinces with high coefficient of variation include rather heterogeneous municipalities, leading to blurred causality links between the drivers under scrutiny and building permits. We try to account for this potential issue in two alternative ways: first, we compute the average coefficients of variation for low, medium and high exposure and exclude, for each of the three levels of risk, those provinces that lie above the 95 percentile. Second, we take the average of the three measures of heterogeneity and exclude those provinces that lie above the 90 percentile of average heterogeneity in risk exposure. Results are reported, respectively, in Tables 7 and 8 and are in line with our baseline results reported in Table 4 in terms of sign, magnitude of coefficients and statistical significance.

Table 7 Robustness check: exclusion of provinces with very heterogeneous municipalities (above 90 percentile) in terms of within-province coefficients of variation of either low, medium or high risk
Table 8 Robustness check: exclusion of provinces with very heterogeneous municipalities (above 95 percentile) in terms of within-province average coefficient of variation of low, medium and high risk

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

D’Amato, A., Marin, G. & Rampa, A. Environmental Disasters and Electoral Cycle: An Empirical Analysis on Floods and Landslides in Italy. Environ Resource Econ 74, 625–651 (2019). https://doi.org/10.1007/s10640-019-00338-7

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10640-019-00338-7

Keywords

Navigation