Target Hardening and Non-State Armed Groups’ Target Selection: Evidence from India

ABSTRACTThis study explores the variation in the non-state armed group (NSAGs)'s behavior concerning target selection. Scholars of transnational terrorism have investigated transnational NSAGs' target selection. However, we are still missing out on the most common form of terrorism, terrorism perpetrated by domestic NSAGs involved in civil conflicts. This paper’s novel contribution is to the understanding of domestic NSAGs’ strategic logic. I argue that hardening makes soft targets, including civilians, attractive targets when hard targets are no longer attractive. NSAGs tactically adapt to hardening by switching to soft targets or by displacing attacks to adjacent locations within their home country. The empirical results from data on relevant state-group dyads in India between 2004–2016 show that domestic NSAGs (1) switch to soft targets when faced with hardening, (2) less frequently target soft targets when more of their attacks against hard targets have been logistically successful, and (3) commit more attacks in their primary area of operation when more of their attacks in that location have been logistically successful. These findings emphasize a variety of ways through which domestic NSAGs adapt their tactics and underscore potential costs for target hardening.KEYWORDS: Hardeningtarget selectionnon-state armed groupsterrorism in civil conflictstargeting of civiliansdomestic terrorism Disclosure statementNo potential conflict of interest was reported by the author(s).Supplementary materialSupplemental data for this article can be accessed online at https://doi.org/10.1080/09546553.2023.2252917.Notes1. Walter Enders and Todd Sandler, “What Do We Know about the Substitution Effect in Transnational Terrorism?” in Research on Terrorism: Trends, Achievements, and Failures, ed. Andrew Silke (Routledge, 2004).2. Patrick T. Brandt and Todd Sandler, “What Do Transnational Terrorists Target? Has It Changed? Are We Safer?” The Journal of Conflict Resolution 54, no. 2 (2010): 214–36; Walter Enders and Todd Sandler, “Distribution of Transnational Terrorism among Countries by Income Class and Geography after 9/11,” International Studies Quarterly 50, no. 2 (2006): 367–93.3. Joseph K. Young and Michael G. Findley, “Promise and Pitfalls of Terrorism Research,” International Studies Review 13, no. 3 (2011): 411–31.4. Brandt and Sandler, “What Do Transnational Terrorists Target?”5. START, “National Consortium for the Study of Terrorism, and Responses to Terrorism,” Global Terrorism Database, 2018, https://www.start.umd.edu/gtd/ (accessed April 27, 2021).6. Ibid.7. The GTD defines a terrorist attack as “the threatened or actual use of illegal force and violence by a non-state actor to attain a political, economic, religious, or social goal through fear, coercion, or intimidation” (10). To be included in the GTD, an incident must (1) be intentional, (2) entail some level of violence, and (3) be perpetrated by a non-state actor. In addition to these three criteria, at least two of the following three criteria must be present: “the act must be aimed at attaining a political, economic, religious, or social goal,” “there must be evidence of an intention to coerce, intimidate, or convey some other message to a larger audience (or audiences) than the immediate victims,” and “the action must be outside the context of legitimate warfare activities” (11). If the incident meets the first two criteria, it is included in the GTD even if it is within the context of legitimate warfare activities. Thus, GTD includes some attacks that target combatants.8. Todd Sandler, “The Analytical Study of Terrorism: Taking Stock,” Journal of Peace Research 51, no. 2 (2014): 257–71; Sebastian Jackle and Marcel Baumann, “‘New Terrorism’ = Higher Brutality? An Empirical Test of the ‘Brutalization Thesis,’” Terrorism and Political Violence 29, no. 5 (2017): 875–901; Victor Asal and Justin V. Hastings, “When Terrorism Goes to Sea: Terrorist Organizations and the Move to Maritime Targets,” Terrorism and Political Violence 27, no. 4 (2015): 722–40.9. See note 4 above.10. Khusrav Gaibulloev, Todd Sandler, and Charlinda Santifort, “Assessing the Evolving Threat of Terrorism,” Global Policy 3, no. 2 (2012): 135–44.11. Asal and Hastings, “When Terrorism Goes to Sea”; Brandt and Sandler, “What Do Transnational Terrorists Target?”12. Walter Enders and Todd Sandler, “The Effectiveness of Antiterrorism Policies: A Vector-Autoregression- Intervention Analysis,” The American Political Science Review 87, no. 4 (1993): 829–44; Todd Sandler and Walter Enders, “An Economic Perspective on Transnational Terrorism,” European Journal of Political Economy 20, no. 2 (2004): 301–16; Todd Sandler, “Collective Action and Transnational Terrorism,” The World Economy 26, no. 6 (2003): 779–802; Daniel G. Arce and Todd Sandler, “Counterterrorism: A Game-Theoretic Analysis,” Journal of Conflict Resolution 49, no. 2 (2005): 183–200.13. Enders and Sandler, ”What Do We Know about the Substitution Effect in Transnational Terrorism?”14. Eli Berman and David D Laitin, “Religion, Terrorism and Public Goods: Testing the Club Model,” Journal of Public Economics 92, no. 10 (2008): 1942–67; James A. Piazza, “Suicide Attacks and Hard Targets: An Empirical Examination,” Defence and Peace Economics 31, no. 2 (2018): 142–159.15. Enders and Sandler, “Distribution of Transnational Terrorism among Countries by Income Class and Geography after 9/11”; Arce and Sandler, “Counterterrorism”; Gaibulloev, Sandler and Santifort, ”Assessing the Evolving Threat of Terrorism.”16. Berman and Laitin, “Religion, Terrorism and Public Goods.”17. Justin V. Hastings and Ryan J. Chan, “Target Hardening and Terrorist Signaling: The Case of Aviation Security,” Terrorism and Political Violence 25, no. 5 (2013): 777–97.18. Brandt and Sandler, “What Do Transnational Terrorists Target?”; Enders and Sandler, “What Do We Know about the Substitution Effect in Transnational Terrorism?”19. Charlinda Santifort, Todd Sandler and Patrick T. Brandt, “Terrorist Attack and Target Diversity: Changepoints and their Drivers,” Journal of Peace Research 50, no. 1 (2013): 75–90.20. Enders and Sandler, “Distribution of Transnational Terrorism among Countries by Income Class and Geography after 9/11”; Brandt and Sandler, “What Do Transnational Terrorists Target?”; Brandt and Sandler, “What Do Transnational Terrorists Target?”21. Gaibulloev, Sandler and Santifort, “Assessing the Evolving Threat of Terrorism”; Santifort, Sandler and Brandt, “Terrorist Attack and Target Diversity.”22. Gaibulloev, Sandler and Santifort, “Assessing the Evolving Threat of Terrorism.”23. William M. Landes, “An Economic Study of U. S. Aircraft Hijacking, 1961–1976,” The Journal of Law and Economics 21, no. 1 (1978): 1–31; Enders and Sandler, “The Effectiveness of Antiterrorism Policies”; Enders and Sandler, ”What Do We Know about the Substitution Effect in Transnational Terrorism?”24. See note 16 above.25. Piazza, “Suicide Attacks and Hard Targets.”26. Max Abrahms and Justin Conrad, “The Strategic Logic of Credit Claiming: A New Theory for Anonymous Terrorist Attacks,” Security Studies 26, no. 2 (2017): 279–304.27. Justin George, “State Failure and Transnational Terrorism: An Empirical Analysis,” Journal of Conflict Resolution 62, no. 3 (2018): 471–95.28. Ilayda Onder, “Signaling Resolve Through Credit-Claiming,” International Interactions (2023), doi:10.1080/03050629.2023.2216352.29. Veronica Persson, “Framing Mediated Terrorism Before and After 9/11: A Comparative Study of ‘Framing’ Kenya and Tanzania in 1998 and Madrid 2004 in the Swedish Broadsheet of Dagens Nyheter” (Master’s Thesis, Stockholm University, 2004).30. Todd Sandler and Daniel G. Arce M., “Terrorism & Game Theory,” Simulation & Gaming 34, no. 3 (2003): 319–37; Daniel G. Arce and Todd Sandler, “Terrorist Signaling and the Value of Intelligence,” British Journal of Political Science 37, no. 4 (2007): 573–86.31. Abrahms and Conrad, “The Strategic Logic of Credit Claiming.”32. Carly Wayne, “Terrified or Enraged? Emotional Microfoundations of Public Counterterror Attitudes,” Working Paper, (2022), https://www.carlywayne.com/terrorism-counterterrorism (accessed May 22, 2023).33. Sandler and Arce M., “Terrorism & Game Theory.”34. Robert Powell, “Defending against Terrorist Attacks with Limited Resources,” The American Political Science Review 101, no. 3 (2007): 527–41.35. Bruce Hoffman and Gordon H. McCormick, “Terrorism, Signaling, and Suicide Attack,” Studies in Conflict & Terrorism 27, no. 4 (2004): 243–81.36. Bruce Hoffman, Inside Terrorism (Columbia University Press, 2006); Jackle and Baumann, “‘New Terrorism’ = Higher Brutality?”37. Seung-Whan Choi and James A. Piazza, “Foreign Military Interventions and Suicide Attacks,” Journal of Conflict Resolution 61, no. 2 (2015): 271–97.38. Michael Townsley, Shane D. Johnson, and Jerry H. Ratcliffe, “Space Time Dynamics of Insurgent Activity in Iraq,” Security Journal 21 (2008): 139–46.39. Peter Baudains, Jyoti Belur, Alex Braithwaite, Elio Marchione and Shane D. Johnson, “The Exacerbating Effect of Police Presence: A Multivariate Point Process Analysis of the Naxal Conflict,” Political Geography 68 (2019): 12–22.40. Sandler and Arce M., “Terrorism & Game Theory”; Nancy A. Morris, “Target Suitability and Terrorism Events at Places Terrorism Target Suitability: Special Essay,” Criminology Public Policy 14 (2015): 417.41. Todd Sandler and Harvey E. Lapan, “The Calculus of Dissent: An Analysis of Terrorists’ Choice of Targets,” Synthese 76, no. 2 (1988): 245–614.42. Arie Perliger, Badi Hasisi, and Ami Pedahzur, “Policing Terrorism in Israel,” Criminal Justice and Behavior 36, no. 12 (2009): 1279–304.43. Figures showing the annual patterns of soft target selection across states are in Online Appendix 5.44. Figures showing the annual patterns of hardening across states are included in Online Appendix 5.45. START, “National Consortium for the Study of Terrorism, and Responses to Terrorism”46. For the purposes of this study, I only rely on domestic NSAGs in India. GTD’s INT_ANY indicator includes information on whether an attack was a domestic or transnational attack.47. Target types are aggregated using GTD’s targtype1 indicator. Incidents are considered attacks against soft targets if targetype1 is one of the following GTD categories: business, abortion-related, educational institution, food supply, journalists, maritime, NGO, other, private citizens, religious figures, telecommunication, tourists, transportation, utilities.48. Incidents are considered attacks against hard targets if targetype1 is one of the following GTD categories: government (general), police, military, airports, government (diplomatic).49. Figures showing the distribution of attacks across all target types are included in Online Appendix 5.50. Telangana is included in the dataset for only four years, because it was separated from Andhra Pradesh in 2014.51. Relevant state-state dyads are pairs of contiguous states or in which at least one of the states is a major power. For more on relevant-dyads approach, See Zeev Maoz and Bruce Russett, “Normative and Structural Causes of Democratic Peace, 1946–1986,” The American Political Science Review 87, no. 3 (1993): 624–38; Douglas Lemke and William Reed, “The Relevance of Politically Relevant Dyads,” Journal of Conflict Resolution 45, no. 1 (2001): 126–44.52. The coding rules for the groups’ primary area of operation are included in Online Appendix 3.53. To calculate the ratio of the number of attacks against soft targets to the total number of attacks, I first aggregated the number of attacks against soft targets using the GTD’s provstate, group_txt, iyear, and targtype1 indicators. Then, I divided the number of attacks against soft targets by the total number of attacks + 0.1. 0.1 is added because the total number of attacks may take the value of zero.54. Assam Police, Govt. of Assam. Special Branch. Government of Assam, https://police.assam.gov.in/portlet-innerpage/special-branch (accessed May 18, 2021).55. Another data limitation in target hardening is the lack of reliable data on the security expenditures of multinational corporations. MNCs operating in conflict zones may devote additional resources to bolster their security. These efforts are not captured by the police data used in this paper. However, this is unlikely to be a major problem. Most counterinsurgency and counterterrorism policing are not private in most countries but rather overseen by governments.56. Baudains, Belur, Braithwaite, Marchione and Johnson, “The Exacerbating Effect of Police Presence.”57. The success of the attacks is not judged in terms of the broader goals of the perpetrator group. Instead, the GTD’s key criterion is whether or not the attack took place. For example, a bombing is considered successful if the bomb exploded.58. Victor Asal and R. Karl Rethemeyer, “The Nature of the Beast: Organizational Structures and the Lethality of Terrorist Attacks,” The Journal of Politics 70, no. 2 (2008): 437–49; “Mapping Militants Project.” n.d. Stanford University.59. NSAGs that can frequently utilize suicide tactics may be more likely to attack hard targets.60. I use COIN Casualties as a proxy for Indian states’ militarization against NSAGs. Militarization of civil conflicts can tip the balance of power against NSAGs, thereby altering NSAGs’ calculations regarding costs and payoffs of attacking soft targets.61. Stronger groups involved in intense conflicts could utilize the resources required to launch successful attacks against hard targets.62. In more ethnically fragmented states, ethno-nationalist groups may be tempted to target members of other ethnic groups.63. It is suggested that Heckman models should include at least one variable in the selection equation that does not appear in the outcome equation (See Anne E. Sartori, “An Estimator for Some Binary-Outcome Selection Models Without Exclusion Restrictions,” Political Analysis 11, no. 2 (2003): 111–38). Such variables are known as exclusion restrictions. I picked State population as my exclusion restriction because it likely influences the probability that a state-group dyad experiences at least one attack in a given year, as most insurgencies are contained in India’s relatively sparsely populated rural parts. However, I do not expect the state population to influence NSAGs’ strategic logic significantly; thus, the population should not predict the prevalence of attacks against soft targets. One counterargument that can be made against using population as an exclusion restriction is that population captures the number of civilians, which is a soft target. This is unlikely to be a problem in India for three reasons. First, it should be noted that soft targets are always plentiful, even in sparsely populated areas. Secondly, the intuitive relationship between population and abundance of soft targets at least partially depends on urbanization levels since populations are thought to be concentrated in urbanized areas that host many soft targets. However, in India, state population and urbanization are not closely related. For example, according to 2011 census figures, state population and the rate of urbanization (See MOSPI. “2011 Census from HIS.” Ministry of Statistics and Program Implementation) are negatively correlated (−0.21, p = .26). The most populated states, such as Uttar Pradesh, Maharashtra, Bihar, and West Bengal, had low urbanization rates (22 percent, 45 percent, 11 percent, and 32 percent, respectively). In contrast, the most urbanized states, such as Delhi and Chandigarh, were not as populated (Delhi’s population was 16,787,000, which corresponds to 1.4 percent of the total country population, and Chandigarh’s was 1,161,000, which corresponds to 0.1 percent of the country population). Finally, in India, the state population seems to be correlated with the abundance of hard targets rather than the abundance of soft targets. My dataset shows a strong positive correlation between state population and the number of police stations in a given state (0.84, p = .000).64. Civilian targets are a subset of soft targets. While soft targets coding includes all incidents that targeted non-hard targets, civilian targets coding only includes incidents whose target type (GTD’s targetype1 indicator) is private citizens & property. For example, an attack that targeted a school building at night while students, teachers, or staff were not inside the building is not included in the civilian coding.65. Before performing log transformation of the dependent variable, a constant value of 1 was added to each observation in order to handle zero values.Additional informationNotes on contributorsIlayda B. OnderIlayda B. Onder is a Ph.D. Candidate in Political Science at The Pennsylvania State University and a Minerva Peace and Security Scholar Fellow at the United States Institute of Peace (USIP). She studies political violence, rebellion, and terrorism with a particular focus on inter-group relations in multiparty conflicts, armed groups’ interactions with civilians, and their non-violent communication strategies.