GTD in Research

Using data that combines information from the Federal Aviation Administration, the RAND Corporation and the GTD, we are able to examine trends in 1,101 attempted aerial hijackings that occurred around the world from 1931 to 2003. We have especially complete information for 828 hijackings that occurred before 1986. Using a rational choice theoretical framework, we use continuous-time survival analysis to estimate the impact of several major counterhijacking interventions on the hazard of differently motivated hijacking attempts and logistic regression analysis to model the predictors of successful hijackings. Some of these interventions use certainty-based strategies of target hardening to reduce the perceived likelihood of success. Others focus on raising the perceived costs of hijacking by increasing the severity of punishment. We also assess which strategies were most effective in deterring hijackers whose major purpose was related to terrorism. We found support for the conclusion that new hijacking attempts were less likely to be undertaken when the certainty of apprehension was increased through metal detectors and law enforcement at passenger checkpoints. We also found that fewer hijackers attempted to divert airliners to Cuba once that country made it a crime to hijack flights. Our results support the contagion view that hijacking rates significantly increase after a series of hijackings closely clustered in time—but only when these attempts were successful. Finally, we found that the policy interventions examined here significantly decreased the likelihood of nonterrorist but not that of terrorist hijackings. Journal Website

Despite the growth in research examining direct economic impacts of terrorism, the indirect impact of terrorism on the stability of local economies has generally been overlooked. Using panel data regression models and the GTD, we examine the impact of terrorism on employment and business outcomes in Italy from 1985 to 1997. We find that terrorist attacks reduce the number of firms and employment in the year following an attack. By disaggregating net outcomes into their component gross flows, we also find that these impacts are primarily attributable to reduced business formations and expansions. Journal Website

In this paper we examine the trajectories of two Armenian terrorist groups: the Armenian Secret Army for the Liberation of Armenia (ASALA) and the Justice Commandos of the Armenian Genocide (JCAG). Both groups began in the mid-1970s and by the early 1980s had become extremely active. However, shortly afterwards, attacks and fatalities attributed to ASALA and JCAG plummeted and by 1988 both groups had effectively disintegrated. The pivotal historical event in our analysis is an especially brutal attack on Paris's Orly Airport in 1983 which we believe undermined the legitimacy of ASALA among its supporters in the Armenian Diaspora and in the West. We use data from the Global Terrorism Database (GTD) from 1975 to 1988 as well as extensive qualitative evidence to examine these issues. Based on Cox proportional hazard models, we find that both total ASALA attacks and ASALA attacks on non Turkish targets significantly increased until the Orly incident, but significantly decline thereafter. Although JCAG was not involved in the Orly bombing and in general had a much more disciplined approach, JCAG attacks also declined rapidly following Orly. The results suggest that when a terrorist organization depends heavily on a Diaspora, over-reaching in terrorist targeting offers a strong opening for discrediting terrorism as a tactic, even discrediting terrorists who have not over-reached. Journal Website

The purpose of this analysis was to provide an exploratory look at a recently computerized database, examining the interplay between nonstate actors, terrorism, and WMD. In this brief overview, we have concentrated on the following risk factors: (1) previous use of WMD, (2) a history of willingness to launch attacks outside of the country of origin, (3) willingness to kill large numbers of people, and (4) attempts to achieve maximum lethality. Along the way, we have identified certain groups that fill these criteria and thus will be candidates for mass impact terrorism in the future. In short, although making no statistically formulated arguments, the data discussed here nonetheless serve as a point of departure for a better understanding of a host of issues of concern in both the academic and policy fields. Journal Website

The existence of violent crime hot spots, both in neighborhoods and countries, which remain relatively stable over time, has been fairly well-established in the criminological literature. In this chapter, we ask if terrorism incidents are also concentrated in a relatively few number of countries and whether that concentration has remained stable over time. We use the Global Terrorism Database which includes 70,000 domestic and international terrorism incidents from 1970 to 1997. We use semi-parametric trajectory analysis to estimate and describe long term patterns of terrorism activity for all countries during the time span. In addition, we classify countries with similar patterns into groups which reflect the overall pattern of terrorism. Our results indicate that there are four trajectory groups in the data, each of which represents a distinct pattern of terrorism activity over time. Although the vast majority of countries experience relatively few numbers of terrorist incidents over time, a small percentage of countries account for the majority of terrorist incidents in our data. Moreover, the clustering of incidents at the country level remains stable over time. For example, countries in trajectory group four represent only 8% of countries in the data, however they accounted for 67% of all incidents between 1970 and 1997. We conclude that there is a substantial amount of concentration of terrorist incidents within particular countries, and this concentration is stable over time. This suggests that the policing of terrorism should not be conducted in a random manner; rather resources for combating terrorism ought to be concentrated in the places with the most terrorism. Book Website

Despite growing international concern about terrorism, until recently, very little was known about world wide risk patterns for terrorist attacks. In this paper we are especially interested in determining the extent to which terrorism is concentrated at the country level over time and whether different measures of terrorism (total, attributed and fatal attacks) yield similar results. Traditional sources of crime data—official police records and victimization and self report crime surveys—have not included terrorism. In response, there has been growing interest in terrorist event data bases. In this research we report on a data base formed by merging the Global Terrorism Database maintained by START with the RAND-MIPT data base. We use semi-parametric group-based trajectory analysis to analyze 73,961 attacks in 207 countries from 1970 to 2006. Our results confirm that terrorism, like more common crime, is highly concentrated at the country level and these concentrations are fairly stable over time. Ten countries account for 38 percent of all known terrorist attacks since 1970; 32 countries account for more than three-quarters of all attacks. The trajectory analysis also reveals a rapidly rising new terrorist threat concentrated among countries from South Asia, South East Asia and Eastern Europe.

Compared to most types of criminal violence, terrorism poses special data-collection challenges. In response, there has been growing interest in open-source terrorist event data bases. One of the major problems with these data bases in the past is that they have been limited to international events—those involving a national or group of nationals from one country attacking targets physically located in another country. Past research shows that domestic incidents greatly outnumber international incidents. In this paper we describe a previously unavailable open source data base that includes some 70,000 domestic and international incidents since 1970. We began the Global Terrorism Database (GTD) by computerizing data originally collected by the Pinkerton Global Intelligence Service (PGIS). Following computerization, our research team has been working for the past two years to validate and extend the data to real time. In this paper, we describe our data collection efforts, the strengths and weaknesses of open source data in general and the GTD in particular, and provide descriptive statistics on the contents of this new resource. Journal Website

As international concern about terrorism has grown, researchers and policymakers have increasingly sought to understand terrorism by looking at the social, economic, and political characteristics of countries. Lafree, Dugan, and Fahey examine connections between a newly available measure of terrorist attacks-the Global Terrorism Database (GTD) covering the period between 1970 and 1997-and state failure, defined by the Political Instability Task Force as including "civil conflicts, political crises, and massive human rights violations that are typically associated with state breakdown." Book Website

Terrorism is a form of crime. Yet compared to most types of crime, terrorism poses unique data collection challenges. As a result, even basic descriptive questions about terrorism have been difficult or impossible to answer: What are the long term trends in terrorist attacks? Is the number of fatalities associated with terrorist attacks increasing over time? What types of attacks are most common? What types of weapons do terrorists use most frequently? How long do terrorist groups last? In this chapter we analyze newly available data from the Global Terrorism Database (GTD) to provide a descriptive account of more than 82,000 domestic and international terrorist attacks that occurred between 1970 and 2004. We provide detailed information on global and country-level terrorism trends, regional characteristics of terrorism, and characteristics of the major groups that have employed terrorist methods. This chapter is meant to provide an overview of the characteristics of global terrorism. We show that many common stereotypes about terrorism receive little support. To place terrorism in a global political context, we include an analysis that links terrorist attacks to political characteristics of nations concentrating especially on level of democratization and state failures. We also examine how terrorism rates compare to more common forms of crime. We conclude with a discussion about important research questions for the future. Book Website

Criminologists since Becarria and Bentham have been concerned with predicting how governmental attempts to maintain lawful behavior affect subsequent rates of criminal violence. In this paper we build on prior research to argue that governmental responses to a specific form of criminal violence—terrorism—may produce both a positive deterrence effect (i.e., reducing future incidence of prohibited behavior) but also a negative backlash effect (i.e., increasing future incidence of prohibited behavior). Deterrence-based models have long dominated both criminal justice and counter terrorist policies on responding to violence. They maintain that an individual's prohibited behavior can be altered by the threat and imposition of punishment. By contrast, research on backlash models applied to either criminal justice or counter terrorist policies are less common and more theoretically scattered. Nevertheless, there is substantial support for such arguments from research on counter terrorism, from criminology research on labeling, legitimacy and defiance, and from the psychological literature on social power and decision making. In this paper we identify six major British strategies aimed at reducing political violence in Northern Ireland from 1969 to 1992 and then use Cox proportional hazard models to estimate the impact of these interventions on the risk of new attacks. In general, we find the strongest support for backlash models. The only support for deterrence models was a military surge called Operation Motorman which showed significant declines in the risk of new attacks. The results underscore the importance of considering the possibility that anti-terrorist interventions might increase as well as decrease subsequent violence. Journal Website

In recent years an increasing number of researchers have observed that there are far fewer studies of how terrorism ends than how it begins. In criminology the issue of how crime ends has been shaped by discussions of desistance, the prolonged or permanent cessation of criminal behavior. We begin this essay with a brief review of research on desistance in criminology, considering first the conceptual challenges of desistance research and then reviewing major theoretical frameworks and empirical findings from criminology that might help inform an understanding of desistance from terrorism. We conclude with a discussion of the implications of desistance research in criminology for research on terrorism at the individual- and group-level and identify several objectives for such a research agenda. Journal Website

While researchers began to assemble open-source terrorism event data bases in the late 1960s, until recently most of these data bases excluded domestic attacks. This is a particularly misleading exclusion for the United States: while the United States is often perceived to be the central target of transnational terrorism, the domestic attacks of the foreign groups targeting the United States are often ignored. We begin this paper with 53 foreign terrorist groups that have been identified by U.S. State Department and other government sources as posing a special threat to the United States. Using the Global Terrorism Database that includes both domestic and transnational terrorist attacks, we examine 16,916 attacks attributed to these groups between 1970 and 2004. We find that just over three percent of attacks by these designated anti-U.S. groups were actually directed at the United States. Moreover, 99 percent of the attacks that targeted the United States did not occur on U.S. soil, but were aimed at U.S. targets in other countries (e.g., embassies or multilateral corporations). We also find that over 90 percent of the non-U.S. attacks were domestic (nationals from one country attacking targets of the same nationality in the same country). We use group-based trajectory analysis to examine the different developmental trajectories of U.S. target and non-U.S. target terrorist strikes and conclude that four trajectories best capture attack patterns for both. These trajectories outline three terrorist waves, occurring in the 1970s, 1980s and the early twenty-first century, as well as a trajectory that does not exhibit wave-like characteristics but instead is characterized by irregular and infrequent attacks. Journal Website

This paper examines spatial and temporal patterns of terrorist attacks in Spanish provinces between 1970 and 1997. Using exploratory spatial data analysis (ESDA), we examine whether terrorist acts are spatially clustered, if these clusters change over time, and if these changes can be attributed to particular diffusion pathways. We find that there are two distinct temporal phases—the first is characterized by increases in the number of attacks accompanied by increases in the dispersion of terrorist attacks across provinces. The second phase is characterized by decreases in overall levels of terrorist activity, but with no accompanying decreases in the dispersion of activity. We also find that increases in province level terrorist activity cannot be attributed to diffusion from nearby provinces, but decreases do diffuse from nearby areas.

Recent increases in terrorist activity around the world have made analyzing and understanding such activities more critical than ever. With the help of organizations such as the National Consortium for the Study of Terrorism and Responses to Terrorism (START), we now have detailed historical information on each terrorist event around the world since 1970. However, due to the size and complexity of the data, identifying terrorists' patterns and trends has been difficult. To better enable investigators in understanding terrorist activities, we propose a visual analytical system that focuses on depicting one of the most fundamental concepts in investigative analysis, the five W's (who, what, where, when, and why). Views in our system are highly correlated, and each represents one of the W's. With this approach, an investigator can interactively explore terrorist activities efficiently and discover reasons of attacks (why) by identifying patterns temporally (when), geo-spatially (where), between multiple terrorist groups (who), and across different methods or modes of attacks (what). By coupling a global perspective with the details gleaned from asking these five questions, the system allows analysts to think both tactically and strategically. Journal Article

A recent explosion of work has occurred seeking to understand under what conditions groups use terror to pursue their goals. Regime type is one condition that is prominently discussed. Democratic regimes have been linked to terrorism for contending reasons with some scholars claiming that constraints on the executive and free civil rights increase the likelihood of these regimes experiencing terror while other scholars claim that democracy promotes nonviolent political participation thus reducing the likelihood of terror. Corroborating evidence has been used for both relationships leading to the following puzzle: why do some democratic regimes seem to foster terrorism while others do not? We offer three contributions to this debate. Beginning with the assumption that terror groups want to change government policy, we argue that the more veto players present in a political system, the more likely the system is to experience deadlock. Given the inability of societal actors to change policies through nonviolent and institutional participation, these regimes will have a greater likelihood of experiencing terror events. Second, we argue that the terrorism data generating process has been misidentified. In short, two types of zeros exist in terrorism data that are different but have been treated the same in previous models. Some states across the entire period of most samples never experience terror while other states only experience terror in a few years and thus have some probability of experiencing terror in the future. We use zero-inflated count models to model this process. Third, we use new data from the Global Terrorism Database (GTD).