The Importance of Governments’ Response to Natural Disasters to Reduce Terrorist Risk

Abstract

The increased incidence of natural disasters may be among the major forces driving crime over the coming century. Indeed, previous research suggests that natural disasters provide the catalyst for acts of terrorism. However, because governments can choose how they respond to natural disasters, they might be able to select actions that would mitigate hostile reactions. In this article, we argue that government responses’ natural disasters could alter the rational motivations for terrorism. If true, then by responding adequately to a disaster, a government can also reduce both political violence risks and public health harms. This study argues that it is the portrayal of government responses to a natural disaster that impacts subsequent terrorism rather than the natural disaster itself. Using data from the Philippines, Turkey, and the United Kingdom between 1987 and 2013, findings suggest that when disasters occur, governments have a unique opportunity to diminish the incentives for terrorism.

Keywords:

• Natural disasters
• terrorism
• disaster response policy
• government actions
• rational choice

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 It should also be noted that terrorist organizations have their own distinct goals that can diverge markedly from simply preserving the well-being of their constituency. For example, despite being called the Kurdistan Workers’ Party (PKK), which implies that its central mission is to support the Kurdish Separatist movement in Turkey (Rodoplu, Arnold, & Ersoy, 2003; Çandar, 2012), the PKK was founded as a Marxist-Leninist group and it also promotes a leftist agenda (Criss, 1995). Indeed, its motivations have been described as being more in line with its founder, Öcalan, than a direct expression of the Kurdish population in Turkey (Özcan, 2006).

2 Other candidates for this study include Nigeria, India, and the Democratic Republic of Congo, as they all experienced geographically diverse terrorism and natural disasters during the period examined. For the purposes of this study, however, Turkey, the Philippines and the UK were selected due to their clear geographic delineation with regard to terrorism. The other nations, however, are ideal contexts to further assess the generalizability of the findings generated by this study.

3 The EM-DAT definition of a natural disaster requires that for an event results in the death of 10 or more people killed, 100 or more people are affected, there is a declaration of a state of emergency, or there is a call for international assistance (CRED, 2010).

4 Groups that were coded as being Islamist in the Philippines include the following: Bangsamoro Islamic Freedom Movement (BIFM), Abu Sayyaf Group (ASG), Moro Islamic Liberation Front (MILF), Moro Ghuraba, Moro National Liberation Front (MNLF), Jemaah Islamiya (JI), Muslim Fundamentalists, Al-Khobar, Muslim Separatists, Muslim Rebels, Muslims, Muslim Militants, Muslim Guerrillas, Tulawie Clan, Former Muslim Guerrillas, Moro National Liberation Front Splinter Group, Bodyguards of Muslim Warlord Governor Tupay Loong, Bangsamoro National Liberation Army, and Bangsamoro Islamic Freedom Movement (BIFM). Groups that were coded as being Kurdish include the following: the Kurdistan Workers’ Party (PKK), Kurdish Militants, Kurdish Separatists, Kurdish Rebels, Kurdish Militants, Kurdish Guerillas, Kurds, Kurdistan Freedom Hawks (TAK), Kurdish Islamic Unity Party, Kaka Tribesmen, and the Hamawand Tribe. Groups that were designated as being Republican include the following: the Irish National Liberation Army (INLA), the Irish Republican Army (IRA), Irish Republican Extremists, the Irish People’s Liberation Organization (IPLO), the Continuity Irish Republican Army (CIRA), the Real Irish Republican Army (RIRA), Óglaigh na hÉireann, Dissident Republicans, and the New Irish Republican Army.

5 Search Syntax for the Philippines: (philippin* or filippin*) and (disaster or earthquake or flood or temperature or epidemic or landslide or avalanche or volcan* or typhoon or tsunami or mudslide)

Search Syntax for Turkey – “turk*” and “disaster” or “earthquake” or “flood” or “temperature” or “epidemic” or “landslide” or “avalanche”

Search Syntax for UK – ("u k" or "united kingdom" or "northern ireland" or "england" or "scotland" or "britain" or "great britain" or "wales" or "belfast" or "cardiff" or "edinburgh" or "london" or "british" or "brittish" or "welsh" or "briton" or "scottish" or "gb" or "irish" or "ulster" or "cymru" or "glasgow" or "aberdeen" or "scots" or "english" or "isle of man" or "guernsey" or "jersey" not "new jersey") and (disaster or earthquake or flood or temperature or epidemic or landslide or avalanche or volcan* or typhoon or tsunami or mudslide)

6 Although this strategy may not a perfectly measure public perceptions, it offers a replicable and reliable way to measure perceptions that is consistent across all three nations.

7 We note here that these control variables only capture yearly variation within each country, as monthly data are unavailable, which will result in downward bias in their standard errors because months in the same year will have the same value. Consequently, significance tests for the control variables may erroneously conclude relevance when there is none. Despite this limitation, we retain these measures for analyses due to their expected relationship to the key independent and dependent variables, which could lead to omitted variable bias if excluded.

8 We run these models separately to avoid multicollinearity that could mask important findings.

9 In some cases, it seems that the government did not physically respond to the natural disaster because its verbal response was their only discovered response to these natural disasters (e.g. “Officials said it was difficult to get a clear picture of the situation because of problems in communicating with the remote area. Anatolian quoted deputy Prime Minister Bulent Ecevit as saying a village in Bingol was totally damaged by the quake.” “There was no word of loss of life, Ecevit said.” Excerpt taken from Quake hits eastern Turkey, 11 injured, lba0000020010925du4d07m93.

10 When robustness checks were conducted using Poisson (in line with Berrebi & Ostwald, 2011), negative binomial, and zero-inflated negative binomial models, the substantive findings held.

11 Subsequent sensitivity analyses were used to examine whether the relationship between natural disasters and terrorism observed in the Philippines may have been contingent on the presence volcanic eruptions, which were absent from the other two nations. The results were nearly identical when volcanic eruptions were removed.

12 The coefficient estimates for the eight full models with statistically significant estimates for the adequacy of the government’s response to natural disasters can be found in Table 3 in the Appendix.

13 This finding for Turkey was robust to a Bonferroni corrected α of 0.004; however, the finding would not be statistically significant for the UK. However, had this study only used one model to test each hypothesis within each nation and region (a lag of one month) following Smith et al., (2016), this would be considered statistically significant.

14 The coefficient estimates for the eight full models with statistically significant estimates for the adequacy of the government’s response to natural disasters can be found in Table 4 in the Appendix.

15 Had we tested the model with two-tails, the adequacy of responses to disasters elsewhere in the Philippines would be positively related to non-Islamic attacks six months later ( EQ \O(β, ^) = 3.239, p = 0.005). However, a closer investigation into this finding suggests that it is driven by the spike in terrorism in 2013, because it disappears when the 2013 data are excluded.

Additional information

Notes on contributors

Daren Fisher

Daren Fisher is an Assistant Professor at The Citadel. He is specialized in empirically testing the predictions of criminological theory to better inform government policies that aim to reduce terrorism using econometric methods and qualitative approaches. Dr. Fisher has published articles in the European Journal of Criminology, Studies in Conflict and Terrorism, Critical Studies on Terrorism, Police Practice and Research, and Women & Criminal Justice. He has also authored book chapters on Sociological and Criminological Explanations of Terrorism (Oxford University Press, eds. Erica Chenoweth, Andreas Gofas, Richard English, and Stathis Kalyvas) and the Emergence of Classical Criminological Theory (Wiley, ed. Ruth A. Triplett).

Laura Dugan

Laura Dugan is a Professor in the Department of Criminology and Criminal Justice at the University of Maryland. Her research examines the predictors and consequences of terrorist violence and the efficacy of violence prevention/intervention policy and practice. She also designs methodological strategies to overcome data limitations inherent in the social sciences. Dr. Dugan is a founding co-principal investigator for the Global Terrorism Database (GTD) and co-principal investigator of the Government Actions in Terrorist Environments (GATE) dataset. Dugan holds a doctorate in Public Policy and Management and a Masters in Statistics from Carnegie Mellon University. She has coauthored Putting Terrorism into Context: Lessons Learned from the World’s Most Comprehensive Terrorism Database, along with more than sixty journal articles and book chapters.