Terrorism and state repression of human rights: A cross-national time-series analysis

Abstract

This study examines the major factors that predict states’ repressive policies, focusing on the relationship between oppositional terror attacks and state repression of core human rights. We rely on a theoretical framework that brings together actor-oriented explanations and socio-cultural approaches. While the former emphasize purposive rational action, international pressures, and domestic threats, the latter focus on the power of ideas and on processes of policy diffusion and cultural norms. Relying on a longitudinal cross-national analysis of panel data for the years 1981–2005, we find substantial evidence for the effects of both actor-oriented measurements and socio-cultural ones. These findings join a growing body of research that emphasizes the importance of the institutional and cultural determinants of states’ counterterrorist policies.

Keywords

Cross-national human rights longitudinal analysis repression spatial diffusion terrorism

Introduction

Numerous studies have examined the relationship between non-state oppositional terrorism (henceforward ‘terrorism’) and state terrorism (henceforward ‘state repression’), mainly in the form of human rights abuses (Shor, 2010). This relationship is by no means a simple one and contentions about its causal direction abound. Many scholars argue that state repression is one of the main precursors of non-state terrorist acts. Indeed, various case studies (e.g. Alexander, 2002; Hewitt, 1984) as well as recent cross-national analyses (Allen and Colley, 2008; Bell et al., 2013; Piazza, 2006; Robison, 2010; Walsh and Piazza, 2010) lend support for this view. The current study focuses on the other direction of the relationship – the effects of terrorism on state repression. In other words, we ask how terrorist acts affect subsequent repression and what other factors may help to explain which states are more likely to repress human rights.

When facing terrorist campaigns, states frequently resort to various repressive measures, often justified as ‘acts of prevention’, ‘acts of deterrence’, or ‘retaliatory measures’ (Shor, 2011).¹ States may direct these measures toward perpetrators of terrorist acts (as in the case of targeted assassinations, although even then non-combatants sometimes get hurt), toward the groups of people accused of supporting their goals, or toward the general civil population of a country. Common forms of state repression, which often follow terrorist acts, are as follows (1) killing or wounding terrorists or civilians; (2) widespread arrests, often without due process and internment without trial; (3) state kidnappings of those accused of carrying or being involved in terrorism; (4) torture; (5) collective punishments, such as closures, destruction of houses or property; and (6) curtailment of civil liberties such as the freedom of movement, speech, or assembly.

Various countries commonly employ one or more of these measures throughout recent history in their fight against terrorism. The first well-researched case in the last two centuries is that of Russia at the end of the 19th century. Following the assassination of Tsar Alexander II, the Russian government began a wave of arbitrary arrests, banishments, large-scale executions of opposition leaders, and the abolishment of liberal reforms (Lacquer, 1999; Lacquer and Alexander, 1987; Scheppele, 2005). One can find more recent examples of human and civil rights curtailment in countries suffering from terror attacks in Cyprus in the late 1950s (Hewitt, 1984), Argentina in the late 1970s (Fontaine, 2002; Sikkink and Booth Walling, 2006), Israel since the 1970s (Gross, 2006; Shor, 2008a, 2008c), the United Kingdom since the 1970s (Bonner, 1993), Sri Lanka since the 1980s (Okudzeto, 2007; Wilkinson, 2001), Russia in the early 2000s (Scheppele, 2005), India since the 1960s (Marwah, 2002), the United States following 9/11 (Greenberg and Dratel, 2005; Murphy, 2006), and many others (Okudzeto, 2007).

However, case studies of states (including democracies) that suffered from terrorist campaigns suggest that countries do not always respond in the same way and with the same degree of repression. First, repression levels often vary within the same country at different periods. Indeed, one can find substantial fluctuations over time in the level of repression of various countries. A few prominent examples from the last three decades are Turkey’s varying responses to the attacks of the Kurdish PKK (Aktan and Koknar, 2002), the ebbs and flows in India’s repressive policies in Kashmir and Jammu (Bhoumik, 2005; Marwah, 2002), Russia’s increasingly repressive response to Chechnyan terrorism during the late 1990s and early 2000s (Scheppele, 2005), and the changing levels of state repression in Sri-Lanka’s policies toward its Tamil minority (Van de Voorde, 2005; Whittaker, 2001).

Furthermore, repression levels following terrorist attacks vary substantially between different countries. While some countries resort to highly repressive policies and engage in mass human rights violations, others adopt much more moderate responses. Examples of the former include the mass murders, torture, imprisonment without charges, and ‘disappearances’ during the ‘Dirty War’ of the late 1970s in Argentina (Fontaine, 2002; Roniger and Sznajder, 1999; Sikkink and Booth Walling, 2006); Russia’s violent attacks in Chechnya during the late 1990s and early 2000s (Kramer, 2005; Scheppele, 2005); and the deterioration in respect for human and civil rights in various provinces of India since the late 1980s (Bhoumik, 2005; Marwah, 2002). Countries that adopted less repressive policies (such as negotiations and structural reforms) following terrorist campaigns include Spain since the 1970s (Brotóns and Espósito, 2002; Jimenez, 1993; Martiet al., 2007), Germany in the 1970s (Heinz, 2007; Pridham, 1981; Sobieck, 1994), and Japan in the 1990s (Itabashi et al., 2002).

This variability suggests a need for study designs that combine longitudinal and cross-national analysis. Only recently researchers began to perform large-scale quantitative cross-national analyses that target the relationship between terrorism and state repression. In particular, three studies examined the effects of terrorism on state repression, all looking at transnational terrorist acts. First, Robison (2010) examined the relationship between international terrorist acts and what he termed ‘state terror’. Robison focused his analysis on 124 developing nations, between the years 1976 and 2002. He found a significant deleterious effect of terrorism on the respect for personal integrity rights, a measure combining states’ practices in regard to extrajudicial killings, disappearances, torture, and political imprisonment. A second study by Piazza and Walsh (2009) analyzed pooled data from 1981 to 2003 for 144 countries, examining not only the composite measure, but also each of the particular practices separately. The authors found that transnational attacks led governments to engage in more extrajudicial killings and disappearances, but had no discernible influence on torture and political imprisonment practices. Finally, Dreher et al. (2010) examined 111 countries over the period 1982–2002. The authors found that transnational terrorism significantly, although not dramatically, diminishes government’s respect for basic human rights. They further reported that terrorism negatively affected respect for all four individual physical rights.

The present analysis follows in the footsteps of these pioneering studies. One of our goals is to examine whether the relationship between terrorism and human rights repression holds with the introduction of additional important explanatory variables, ones which scholars of comparative sociology often emphasize, such as measures for globalization and for spatial diffusion (looking at repression in neighbor states). We further seek to reexamine each of the disaggregated measures of repression (extrajudicial killings, disappearances, torture, and political imprisonment). As the studies by Piazza and Walsh (2009) and by Dreher et al. (2010) report diverging results for the effects of terrorism on some of these measures, it is important to further examine how terror events affect each repressive practice separately.²

Beyond this replication aspect, our study also offers two novel contributions. First, we wish to contrast two major theoretical views in the fields of international relations (IR) and political sociology regarding the determinants of state repression. We argue that terrorist attacks constitute one form of threat to nation states (similar to other threats, such as civil wars and internal dissent), which according to actor-oriented theorists of IR might drive states to respond with aggression. We examine this view in conjunction with socio-cultural approaches, which often attribute states’ repressive policies to the effects of non-governmental pressure groups, isomorphic processes, and cultural traditions.

Our second important contribution is the use of a measure of terrorism that focuses on domestic rather than transnational events. Although instances of domestic terrorism greatly outnumber instances of international terrorism, earlier databases on terrorist events, such as the RAND database, the ITERATE database, and the Patterns of Global Terrorism excluded domestic terrorist events and reported only on international ones. Sanchez-Cuenca and de la Calle (2009), who reviewed the literature on domestic terrorism, suggest that this focus on international terrorism may have affected previous cross-national studies on terrorism, its determinants and its consequences, or at least missed an important part of the phenomenon (see also Asal and Rethemeyer (2008)). In the present study, we use the recently released Global Terrorism Database (GTD), (LaFree and Dugan, 2008a), which includes data on close to 100,000 domestic and international terrorist events around the world since 1970 (for further details on this database see our ‘Data and Measurements’ section). While previous studies already examined domestic terrorism as a measure of robustness (Dreher et al., 2010), the current study focuses on this measure exclusively.

Theoretical frame and research hypotheses

We seek to identify the main factors that help explain variations between countries in their level of respect for human rights following terrorist acts. Two major overarching theoretical approaches provide a useful framework for examining this issue: actor-oriented approaches and socio-cultural approaches. Below is a short review of these approaches and the major theoretical arguments within each of them as they relate to states’ adoption of repressive measures.

Actor oriented approaches

Some of the leading researchers of state repression prioritize explanations that focus on the properties of the actors involved in decision-making. They argue that governments engage in repression when they believe that there is a real threat to their regime and that the benefits of using repressive measures exceed the costs of such measures (Gurr, 1986; Moravcsik, 1995, 2000; Poe et al., 1999; Poe and Tate, 1994). According to Poe et al. (1999), political leaders are rational actors who ‘choose to commit abuses of personal integrity rights because they see these inhumane actions as the most effective means to achieve their ends … When faced with threats, leaders will consider responding to them with repression’ (p. 293). This approach provides an important point of departure for the current analysis. It suggests that since governments often perceive terrorism as a real threat, they will use any measures available to them, including ones that repress human rights, in order to eliminate or reduce future terrorist threats. We may therefore expect the following:

H1. A larger number of terrorist attacks in a given country will increase the likelihood that this country will adopt various repressive policies.

The straightforward utility model is moderated by the recognition of most IR scholars that rational state action is often directed primarily at appeasing powerful international players or local constituencies. According to realist approaches in IR, a strong relationship with a great power should affect a country’s respect for human rights and moderate its violent counterterrorist policies. However, the way in which this moderation occurs is controversial. While some scholars stress the role of pressures from great powers (particularly the United States and West European countries) in states’ compliance with humanitarian norms (Ikenberry and Kupchan, 1990; Krasner, 1993, 1999; Waltz, 1979), others maintain that political interference and pressures from western countries or close relationships with them often push countries, especially in the developing world, to adopt more repressive policies (Chomsky, 1988, 1991; George, 1991; Gill, 2004; Lustick, 2006; Sikkink, 2004). In other words, government which maintain substantial political and economic relations with the United States and are dependent on US aid may receive more military assistance that allows for repressive policies, as well as political support that helps dealing with pressures from human rights organizations and the international community to reduce repressive practices. We therefore suggest the following:

H2. Countries that exhibit higher economic or military dependency on the United States will be more likely to repress core physical rights.

According to the logic of liberal IR approaches (Moravcsik, 1995, 1997; Poe et al., 1999), states choose to react to terrorism with repressive measures because they believe this is the best way to stop or at least reduce these attacks, but also in order to win the support of their local constituency and guard against oppositional forces. One would therefore expect governments that are relatively unstable and unpopular to respond with more repressive measures following terrorist attacks. In addition, this approach emphasizes the importance of other factors that may produce domestic pressures and consequently increase repression. These include, for example, political pressures, intensified by events such as state involvement in wars (including civil wars) and international disputes (Gurr, 1986; Poe and Tate, 1994).

H3. Countries involved in violent disputes (whether international or domestic) and those suffering from political instability and unrest will be more likely to adopt repressive policies.

Socio-institutional and cultural approaches

In contrast to actor-oriented approaches, socio-cultural approaches in IR and political sociology tend to emphasize properties of the social and cultural environment and their influence on decision-making. Below we present three related approaches that fall within this broader tradition. While distinct, each points to the effects of the environment and cultural traditions in which a given government operates and adopts policies.

Constructivist scholars of IR believe that humanitarian norms diffusion is a contested process, which social actors must constantly push in order to preserve its inertia. While some emphasize the role of states in spreading human rights norms (Krasner, 1999; Wendt, 1999), most constructivists emphasize the role of non-state social agents working to diffuse these norms. Central among these agents are non-governmental social movements and organizations, both local and transnational (Gränzer, 1999; Keck and Sikkink, 1998; Lutz and Sikkink, 2000; Smith et al., 1998; Tsutsui, 2004), and legislative institutions and activists (Hafner-Burton and Tsutsui, 2005; Risse et al., 1999; Scheppele, 2005; Shor, 2008a, 2008c; Sikkink and Booth Walling, 2006). We propose the following:

H4. Higher involvement in the global system and various transnational bodies will be associated with a lower likelihood for adopting repressive policies.

Sociological new-institutionalism scholars believe that organizations, including states, are especially likely to emulate behaviors and structural components from their environment when facing complex situations, in which there is a high degree of uncertainty (DiMaggio and Powell, 1983, 1991; Meyer and Rowan, 1977; Meyer and Scott, 1992). According to the logic of world polity theory (Boli and Thomas, 1997; Meyer et al., 1997), when confronted with such uncertainty, states are likely to adopt measures that other states commonly use, leading to a process of behavioral convergence. States may adopt such measures even when it is not clear if they will lead to the reduction of terrorism, simply because ‘something must be done’. Furthermore, neo-institutionalists offer that countries often adopt human rights norms and practices, which may counterbalance repressive measures, not because states are convinced that they are important or moral, but simply in an effort to legitimize state action and reduce future uncertainty (Hafner-Burton and Tsutsui, 2005; Meyer et al., 1997). Hence

H5. Countries would be more likely to adopt repressive policies following the adoption of similar policies by neighboring countries.

Finally, Cultural Explanations to state policies tend to emphasize common traditions, ideologies, and value systems. In explaining state repression, they emphasize dehumanization and the exclusion of certain individuals from the realm of those who deserve rights protection (Huntington, 1996; Kuper, 1981), and the belief in ends that justify the use of repressive means (Cole, 2013; Heinz and Frühling, 1999; Pion-Berlin and Lopez, 1991). States with cultures that value principles such as democracy, liberalism, and individual human rights would be less likely to adopt repressive behaviors. On the other hand, states and governments with value systems that emphasize state power, collective uniformity, nationalism, religiosity, and conservative traditions would be more likely to adopt repressive measures. These latter regimes are often non-democracies. Indeed, former studies on repression emphasize the importance of regime type and the protective effect that democracies have over compliance with human rights norms (e.g. Davenport, 1995, 1996, 2007; Davenport and Armstrong, 2004; Henderson, 1991, 1993; Poe et al., 1999).

H6. Democracies and Commonwealth countries will be less likely to use repression, while Muslim countries will be more likely to use repression.

Data and measurements

Table 1 summarizes the variables examined in the study. It includes data on the measurement of variables, where data came from, the years for which data was available, the number of countries for which data were available, and key descriptive statistics.

Table 1.

Descriptive statistics of study variables, grouped by theoretical approach.

Variables	Variable description	Source	No. of states^a	Years	Mean	SD	Min	Max
Dependent
Repression scale (high = 8)	A composite additive index of physical integrity rights (torture, killings, imprisonment, and disappearances)	Cingranelli and Richards (2010)	195	1980–2009	3.12	2.32	0	8
Extrajudicial killings (high = 2)	Government killings without due process of law	Cingranelli and Richards (2010)	195	1980–2009	0.49	0.50	0	2
Disappearance (high = 2)	People disappeared and not found (obvious political motivation)	Cingranelli and Richards (2010)	195	1980–2009	0.24	0.43	0	2
Torture (high = 2)	Purposeful inflicting of extreme pain by government officials	Cingranelli and Richards (2010)	195	1980–2009	0.58	0.49	0	2
Imprisonment (high = 2)	Political incarceration due to speech, non-violent practices, or membership in a group	Cingranelli and Richards (2010)	195	1980–2009	0.80	0.40	0	2
Independent
Actor oriented approaches
Number of terrorist incidents (ln)	Natural log of the number of attacks by non-state political actors against non-combatants	Global Terrorism Database^b	194	1970–2004	−7.17	6.33	−11.51	6.67
Terror events in neighbor states (ln)	Natural log of yearly number of cases	Global Terrorism Database	194	1970–2004	3.53	1.77	−11.51	7.20
Military assistance priority	The cumulative US security assistance (1980–2004) as a percent of the total US security assistance during these years	Finkel et al. (2005)	172	1980–2004	0.53	3.41	0	38.71
Strikes (ln)	Number of general strikes	Banks (2008)	262	1815–2007	−10.62	3.13	−11.51	2.56
Riots (ln)	Number of riots	Banks (2008)	262	1815–2007	−9.74	4.29	−11.51	4.01
Revolutions (ln)	Number of revolutions	Banks (2008)	262	1815–2007	−9.79	4.14	−11.51	2.20
Demonstrations (ln)	Number of anti-government demonstrations	Banks (2008)	262	1815–2007	−9.26	4.73	−11.51	4.01
Regime stability	A dichotomous measure, where unstable regimes are defined as those that retained power for less than 5 years or are about to be replaced within 2 years	Marshall (2014)	190	1980–2004	0.66	0.47	0	1
Civil war (war = 1)	Dummy variable indicating the presence of a civil war with at least 1000 casualties	The Peace Research Institute Oslo (2009)	189	1816–2006	0.03	0.16	0	1
International conflicts (ln)	The number of international conflicts with at least 1000 casualties	MID3: Ghosn and Bennett (2005)	189	1816–2004	−11.00	2.44	−11.51	1.79
Socio-cultural approaches
Globalization index	A combined index of economic, social, and political globalization	Dreher (2006)	194	1970–2010	45.49	17.75	11.92	92.72
Treaty ratification	Dummy variable noting whether a country has ratified at least one major human rights treaty	Hafner-Burton and Tsutsui (2005)	167	1976–2004	0.62	0.49	0	1
Avg. repression in neighbor states	Average repression in countries that share a border with the core country	Marshall (2014)	190	1980–2004	3.10	1.46	0	6.75
Democracy	General openness of political institutions	Marshall (2014)	190	1980–2004	−0.14	7.53	−10	10
Muslim (Muslim = 1)	Countries with more than 50% of the population Muslim	World Bank (2007)	208	1960–2007	0.25	0.43	0	1
Commonwealth (yes = 1)	Part of the Commonwealth of Nations	Wikipedia (2009)	219	1960–2009	0.25	0.43	0	1
Controls
Population, Total (ln)	Natural log of a country’s population in a given year	World Bank (2007)	208	1960–2005	2.09	9.62	9.62	20.99
GDP per capita (ln)	Natural log of gross domestic product per capita in constant 2000 US dollars	World Bank (2007)	208	1960–2005	7.45	1.53	4.03	10.99
Urbanization	Percent of total population that is urban	World Research Institute (2006)	221	1965–2005	50.22	25.46	2.20	100

Number of countries (including political entities that no longer exist, such as the USSR) for which data were available for at least one country year. Actual analyses were conducted for a smaller number of nations (see Appendix 1).

The Global Terrorism Database (1 and 2; LaFree and Dugan, 2008a, 2008b).

Dependent variable

We obtained data for the dependent variable – state repression of core human rights – for the years 1981–2009 from the Cingranelli–Richards (CIRI) dataset (Cingranelli and Richards, 2010). This database presents quantitative information on government respect for 15 internationally recognized human rights for 195 countries. The information in the database relies both on the annual US Department of State’s Country Reports on Human Rights Practices, and on Amnesty International’s Annual Reports. CIRI is increasingly used both inside and outside academia, and has the advantages of (1) being updated annually, (2) providing separate data on physical integrity rights and on the repression of civil freedoms, and (3) containing data on specific (rather than just aggregated) state practices.³ In order to make the interpretation of the results more intuitive (where a higher rank means more repression), we reversed the original scale or the CIRI measure.⁴ We examine the following four physical integrity rights, as well as the combined scale of all four:

Extrajudicial Killings – killings by government officials without due process of law, including murders by private groups if instigated by the government.

Disappearances – cases in which people disappeared, political motivation appears likely, and the victims were been found.

Torture – purposeful inflicting of extreme pain, whether mental or physical, by government officials or by private individuals at the instigation of government officials.

Political Imprisonment – the incarceration of people by government officials because of: their speech, their non-violent religious practices, or their membership in a group.

Independent variables

We organize the various predictors used in the analyses by the theoretical approach they best represent. We recognize that at least some of these predictors may represent more than just one theoretical approach, and that these approaches are not necessarily mutually exclusive. However, we believe that there is value in trying to assess which theoretical assumptions each of these variables most closely measures. Below we briefly describe the logic that led to the inclusion of each of these variables in the analyses.

1. Measurements of actor-oriented approaches

Terrorism. Actor-oriented instrumental rationality approaches suggest that terrorism should have a direct impact and lead to a substantial increase in repressive policies (H1). Following a growing consensus among social science scholars (e.g. Bergesen and Lizardo, 2004; Ganor, 2005; Goodwin, 2006; Piazza, 2006; Robison, 2007), we define terrorism here as the strategic use of violence by non-state political actors against non-combatants for symbolic purposes, usually with the intention of influencing policies. In order to adhere most closely to this definition, our measure of terrorism excludes events in which the main target of the attack was security personnel or armed forces (which many would define as guerilla attacks).⁵

The main predictor, appearing in all of the equations, is the logged⁶ number of terrorist events in a given country-year.⁷ This measure was constructed from the recently released GTD, collected by the National Consortium for the Study of Terrorism and Responses to Terrorism (START; LaFree and Dugan, 2008a). The GTD includes data on close to 100,000 domestic and international terrorist events around the world since 1970.⁸ In an attempt to capture not only actual levels of terrorism, but also the level of terrorist threats, we also include in all models a measure of the number of terror attacks in neighbor states. This measure accounts for the plausible proposition that countries may react to terrorist acts in neighboring countries, seeking to prevent terrorists and ideas from permeating across borders.

To assess our second hypothesis, regarding the influence of external normative and political pressures by the United States, we use US military assistance as a percent of the total US security assistance for the years 1982–2004.⁹ According to different realist scholars (see above), US influence has the potential to affect state repression in two contrasting ways: either by increasing commitment to human rights principles and ideas (thus decreasing repression) or by influencing states to ignore these principles in the name of real politics and the adoption of policies that serve American interests (thus increasing repression).

Our third hypothesis emphasizes the importance of internal pressures and instability. We use the following measures to assess these conditions: the number of (1) strikes, (2) riots, (3) revolutions, and (4) demonstrations in a given year (all logged); (5) regime stability,¹⁰ (6) the presence of a civil war, and (7) the number of international conflicts (logged). We predict that all these measures will increase state repression of physical integrity rights and civil liberties.

2. Measurements of socio-cultural approaches

Socio-institutional and world polity approaches emphasize processes of globalization and diffusion and the influence of normative international conventions and non-governmental organizations on state policies. We use the globalization index, developed by Dreher (2006), (Dreher et al., 2012), as a measure of globalization and global influence. We expect regimes with higher levels of social, political, and economic integration in the global system to be more exposed to world culture and to political and economic pressures. These regimes should therefore be more likely to adopt international humanitarian principles. Similarly, we predict that the ratification of human rights treaties will decrease repression. To evaluate diffusion effects, we use the average level of repression in neighboring countries in the previous year. If we find that the practices of neighbors substantially affect countries, regardless of other threats and pressures, this would suggest the existence of diffusion effects.

Democracy is perhaps the single most studied predictor of state repression (Davenport and Armstrong, 2004). According to Davenport and Armstrong (2004), when studying state repression, a minimalist definition (although still one that includes procedural as well as behavioral aspects) should be preferred in order to avoid conflating measures of repression and democracy. Following these authors and others (Dreher et al., 2010; Piazza and Walsh, 2009) we also adopt a minimalist measure of democracy – the commonly used Polity IV index, which identifies nations along a scale ranging from −10 (‘strongly autocratic’) to +10 (‘strongly democratic’).¹¹ We expect democracies (in particular liberal democracies) to uphold a set of values and beliefs that should make them more respectful of human rights. A second cultural measure used in the various analyses is the prevalence of Islamic culture in a given country. Samuel Huntington (Huntington, 1996) and other proponents of the cultural (or primordial) approaches (e.g. Barber, 1995; Blankley, 2005) argued that Muslim countries (operationalized here as countries in which over 50% of the population is Muslim) tend to be more fundamentalist and less committed to human rights principles. Finally, some former studies (e.g. Mitchell and McCormick, 1988; Poe et al., 1999) suggested that certain Western attitudes and cultural norms (such as respect for individual rights principles) permeated former British colonies (Commonwealth countries) and assimilated into the local cultures, making them more respectful of human and civil rights.

Controls. We include in all models three control variables: (1) population size, (2) gross domestic product (GDP) per capita (constant 2000 US$; logged), and (3) urbanization (percent urban).Countries with larger populations (e.g. China, India, or Indonesia) may be able to better fend off humanitarian pressures from economic and political powers, and thus adopt more repressive policies following terrorist acts. In addition, a larger population may introduce higher levels of internal pressure and unrest, leading to more repression. Regimes that enjoy higher levels of GDP per capita are likely to have a wider variety of options to fight terrorism, and therefore may adopt less repressive measures in response to terrorist attacks. On the other hand, a poorer regime, which lacks financial means, advanced technological measures, or efficient bureaucracies, may turn to wide repression simply because this is its most available policy. As for the level of urbanization, modernization approaches suggest that a country’s level of urbanization is a good measure of that country’s level of modernization. We would expect more modernized countries, in turn, to adopt advanced technological means and thus obtain a wider and less repressive set of tools with which to fight terrorism. In addition, modernization views argue that urbanization facilitates literacy and the expansion of middle classes, which in turn are likely to lead to political liberalization and the adoption of human rights norms (Brachet-Marquez, 1992; Robinson, 1995).

Samples and analyses

We use pooled cross-national time-series regressions to analyze panel data ranging in years from 1981 to 2005. Our large-scale approach has two important advantages over small-scale comparative case studies research designs. First, the quantitative analysis relies on a much larger sample, which in turn increases our ability to generalize beyond the cases examined in the study (see Shandra et al., 2008, 2009). In fact, since the current study covers most countries of the world, our results enjoy very high external validity. The second advantage of our methods lies in the meticulous nature of the data analysis. We examine states’ actions and terrorist events over a long period, minimizing the danger of focusing too much attention on spectacular or well-known incidents and responses and overstating their importance while neglecting other important occurrences.

Analyses cover all countries for which data are available (see Appendix 1). As is common in cross-national time-series study designs, the unit of analysis is the country-year. One of the main advantages of the pooled time-series method over cross-sectional ordinary least squares (OLS) methods is that it increases the number of observations and thus maximizes the number of degrees of freedom and increases the flexibility of the statistical analysis. The large number of observations also helps in minimizing the influence of outlying country-years. Finally, the pooled analysis allows for variance in the policies of the same country at different points in time, enabling the researcher to address subtle changes over time in the dependent variable (Robison, 2007).

We use ordinal logistic regression analysis for panel data in all of our models, conducted using STATA version 13. We also use a random effects (RE) model; the results of the Hausman test statistic (all non-significant) indicate a lack of systematic differences between the point estimates from either fixed effects (FE) or RE models, suggesting that the latter may be preferred on efficiency grounds as well.¹² RE models also have the advantage of allowing us to capture the effects of theoretically important time-invariant variables, such as US military aid, a country having a Muslim majority, and a county coming from a Commonwealth tradition, which we cannot assess using a FE model.

In order to account for possible global time trends in state repression, we include in all analyses a FEs time measure. We lagged all predictors by 1 year. Recent research designs of this type (e.g. Allen and Colley, 2008; Burgoon, 2006; Robison, 2010; Robison et al., 2006; Walsh and Piazza, 2010) often use a lag of 1 year, as it enables the researcher to better capture causality, by allowing social policies some time to percolate (Burgoon, 2006). However, in order to also capture the longer-term effects of various predictors and policies, we also ran statistical models in which the independent variables were lagged 2–5 years.¹³ Outliers appeared in the analysis, but were not significantly influential. We thus chose to leave them in the data. Pairwise correlation analysis showed that no two variables had a correlation higher than 0.5. Variance inflation factor (VIF) analysis (conducted using OLS) further suggested that multicollinearity is not likely a major problem in our models (no VIF score higher than 5).

One of the major problems that might influence our analysis is reverse causality, leading to endogeneity. In this particular case, one can reasonably suggest that government repression may lead to civil unrest and resentfulness and as a result to non-state violence, including in the form of terrorist acts. Indeed, a number of recent studies (Bell et al., 2013; Piazza, 2006; Robison, 2010; Walsh and Piazza, 2010) suggest that state repression increases terrorism. In other words, state repression may not be the result of terrorism, but might instead be its cause. Similarly, one might suspect that our measure of policy diffusion (neighbors’ repression scores) suffers from reverse causality, as a country’s practices might actually affect the policies of its neighbors rather than the other way around. As one of the results of endogeneity may be an underestimation of the standard errors, we use robust standard errors in all of our analyses (Achen, 2000).

In order to further assess the possibility of reverse causality and potential endogeneity, we also report results that were obtained by employing an instrumental variable regression technique. Following Dreher et al. (2010) and Roodman (2006), we used Arellano–Bond panel-data estimation, with one-step difference generalized method of moments (GMM; see also Arellano and Bover (1995)). In these models, we treat the lagged dependent variable, our measure for terrorism, and our measure for policy diffusion as endogenous and the other independent variables as exogenous. We report results of the Sargan–Hansen test on the validity of the instruments and Arellano–Bond tests of first- and second-order autocorrelation. When using the lag of all variables as instruments, the number of instruments amounts to 171. Therefore, as suggested by Roodman (2006), we collapsed the matrix of instruments, combining them into a single column. This approach reduced the number of instruments to a maximum of 66.

Findings

Table 2 presents the basic findings for the effects of various predictors on state repression of physical integrity rights (extrajudicial killings, disappearances, torture, and political imprisonment). As might be expected, the lagged dependent variable has a robust positive effect on the dependent variable, suggesting substantial inertia in state policies. Our focal independent variable, the coefficient for the log of the number of terror events, is positive and significant in all models except for the last (torture), supporting our first hypothesis that terrorism is associated with increased repression. These findings remained consistent and models were largely unchanged when we examined alternative measures of terrorism.¹⁴ Our findings thus provide support for arguments that following terrorist attacks, states tend to react with repression of core human rights, regardless of other conditions and limitations.

Table 2.

Ordinal logistic regression panel analyses of factors influencing state repression of physical integrity rights, 1981–2005 (all predictors lagged 1 year).

	Cumulative scale	Killings	Disappearances	Imprisonment	Torture
Lagged dependent	0.73*** (14.32)	1.26*** (9.44)	1.42*** (9.30)	1.49*** (11.71)	1.21*** (10.94)
Actor-oriented approaches
Terror events (ln)	0.03*** (3.17)	0.05*** (3.95)	0.05*** (3.54)	0.05*** (4.51)	0.01 (1.01)
Terror events in neighbor states (ln)	0.01 (0.31)	−0.00 (−0.03)	0.06 (0.95)	0.03 (0.65)	0.06 (1.17)
US military aid	0.05*** (2.71)	0.03 (1.35)	0.01 (0.66)	0.06*** (5.79)	0.04*** (3.10)
Strikes (ln)	0.23 (1.47)	0.21 (1.09)	−0.17 (−0.84)	0.18 (0.82)	0.19 (0.95)
Riots (ln)	0.10 (0.60)	0.04 (0.22)	0.43** (2.07)	−0.15 (−0.88)	0.16 (0.79)
Revolutions (ln)	0.65*** (4.83)	0.75*** (4.46)	0.58*** (3.34)	0.42*** (2.68)	0.61*** (3.48)
Demonstrations (ln)	0.12 (1.04)	0.10 (0.74)	−0.35** (−1.99)	0.43*** (2.68)	0.26* (1.76)
Stable regime	0.06 (0.51)	−0.02 (−0.10)	−0.04 (−0.23)	0.29** (2.12)	0.01 (0.05)
Civil war	0.92*** (2.81)	0.86** (2.35)	1.00*** (3.37)	0.70* (1.85)	0.45 (1.48)
International conflicts (ln)	−0.01 (−0.48)	−0.01 (−0.42)	0.03 (1.67)	−0.01 (−0.33)	−0.03* (−1.78)
Socio-cultural approaches
Globalization index	−0.03*** (−3.65)	−0.04*** (−2.71)	−0.05*** (−3.39)	−0.03*** (−2.70)	−0.02** (−2.06)
Human rights treaties ratification	0.08 (0.82)	0.13 (1.08)	0.08 (0.59)	0.05 (0.49)	0.08 (0.71)
Avg. repression in neighbor states	0.16** (2.42)	0.14 *(1.72)	0.12 (1.35)	0.19** (2.42)	0.08 (1.05)
Muslim country	−0.16 (−0.86)	−0.44* (−1.70)	−0.55** (−2.28)	0.38* (1.93)	−0.14 (−0.53)
Commonwealth country	0.14 (0.84)	0.45** (2.04)	−0.12 (−0.51)	0.13 (0.66)	0.21 (0.90)
Democracy	−0.08*** (−5.53)	−0.05*** (−2.66)	−0.03 (−1.62)	−0.12*** (−7.97)	−0.07*** (−4.23)
Controls
Population (ln)	0.37*** (6.71)	0.37*** (5.07)	0.20*** (2.68)	0.33*** (4.34)	0.41*** (5.93)
GDP per capita (ln)	−0.27** (−2.13)	−0.33* (−1.78)	−0.05 (−0.33)	−0.18 (−1.15)	−0.45*** (−3.04)
Urbanization	0.01** (2.49)	0.01* (1.67)	0.01 (1.12)	0.01** (2.17)	0.02*** (2.92)
Number of countries	140	140	140	140	140
Observations	2208	2208	2208	2208	2208

The cumulative scale in the first model is a composite index referring to the violation of physical integrity rights. We reversed original Cingranelli–Richards 9-point scale so that a score of 0 represents low levels of repression while a score of 8 represents high repression levels. The disaggregated measurements of repression (killings, disappearances, imprisonment, and torture) were also reversed, with higher scores now representing higher repression levels. We introduce a fixed effects control for year in all models to control for time trends in repression.

Robust t statistics in brackets. *p < .10; **p < .05; ***p < .01.

It is interesting to compare these findings to those of previous studies that examined the relationship between terrorism and respect for specific human rights using a measure of transnational terrorism. Piazza and Walsh (2009) reported that transnational attacks increased extrajudicial killings and disappearances, but had no discernible influence on torture and political imprisonment practices, Dreher et al. (2010), however, found that terrorism significantly, although not dramatically, diminishes government’s respect for all four individual dimensions of physical rights integrity. Our analysis of domestic terrorism provides further support for this latter finding, suggesting that terrorism is indeed detrimental for most core human rights, excluding torture.

Other measurements for actor-oriented approaches were also mostly positive and significant, providing support for this theoretical approach. First, supporting our second hypothesis, the regression coefficients for military US aid was positive and significant in Model 1 (the cumulative scale), lending support to the claims of scholars who argue that higher levels of economic and military dependence on the United States tend to have deleterious effects in terms of countries’ respect for human rights. Our findings show that the two practices that especially flourish under such conditions are imprisonment and torture, perhaps because these are practices that are not as overt (compared to extrajudicial killings) and governments are better able to conceal them from the public eye and the media. Thus, they may be less visible to Western publics and human rights organizations and result in less critique on the US government for turning a blind eye to (or even assisting) violating governments.

A more diverse picture emerges when looking at the measures for liberal IR approaches. While the coefficients for strikes and riots were mostly non-significant, the number of revolutions in a country is a robust predictor of all four measures of repression. Demonstrations also increase the likelihood of imprisonment and torture, although their effect on disappearances is reversed (accounting perhaps for the non-significant effect for the cumulative scale). Consistent with the results of previous studies (e.g. Dreher et al., 2010; Piazza and Walsh, 2009), we also found that civil war is consistently a significant predictor of the aggregated measure for state repression, as well as the repression of most core right individually (except for torture).

Taken together, these results demonstrate that the presence of internal disquiet and violence is very important in predicting repressive policies, and that high levels of domestic unrest are harmful to human rights practices regardless of the levels of terrorism that a country experiences. As noted above, terrorist attacks often lead to political unrest and internal dissent, which undermine governments and push them towards using repression in the hopes of regaining control and suppressing the dissent. Furthermore, as the latest events in Algeria, Egypt, and Syria demonstrate, even non-violent political dissent may push governments (in particular non-democratic ones) to use repressive measures. While many of the measures for domestic political unrest are significant, the measure for international disputes is not, except for a curious negative effect on torture (suggesting that international conflicts might actually reduce torture practices rather than increasing them).

Socio-cultural approaches also receive support in our results. First, in line with our fourth hypothesis, Table 2 shows that a country’s connection to the global system, as measured by its globalization index, is a significant predictor of lower levels of overall repression, as well as the individual repression measures. A possible explanation for this finding is that in countries with more economic, political, and social openness and dependence, regimes may feel more compelled to abide by global and international human rights norms. This may be because they have more to lose by violating these norms, but also because the norms are likely to permeate local practices as their moral power increasingly affects people and institutions.

The ratification of global human rights treaties, on the other hand, has no statistically significant effect, but the direction of the coefficients suggests that it might actually be somewhat harmful to human rights practices. This finding is consistent with the practice which Hafner-Burton and Tsutsui (2005) refer to as ‘empty promises’ – the fact that governments often ratify human rights treaties as a matter of window dressing and at least in the short run actually exacerbate repressive human rights practices (see also Vreeland, 2008). These results suggest that the ability of transnational agents and legislation to reduce serious violations of basic human rights is limited. Hence, as I argued in previous research (Shor, 2008a), the power of norms and ideas to change repressive policies, especially in the face of terrorist threats, should not be overstated.

Our findings regarding regional effects provide more conclusive support for socio-institutional views. Table 2 shows that regimes in countries with neighbors in which the average repression levels had been higher in the previous year were more likely to subsequently repress physical integrity rights (as measured by the cumulative scale). This clearly entails a process of policy diffusion between neighboring countries, where events and policies that are not directly related to the country in question increase the chances that this country will adopt repressive policies, in particular, extrajudicial killings and political imprisonment. This process is in line with neo-institutional and world polity explanations that emphasize processes of policy diffusion, especially when actors face high levels of uncertainty.

Consistent with the results of previous studies, the level of democracy was a robust and powerful predictor of state repression. Regimes that are more democratic were less likely to engage in serious violations of core human rights in comparison to their less democratic counterparts. However, other measures of cultural traditions, such as British cultural tradition (a Commonwealth of Nations country) or the dominance of Muslim culture in the country produced mixed results. While extrajudicial killings and disappearances are actually less likely in Muslim countries, the practice of political imprisonment is more likely. These findings put into question the monolithic assumption that Muslim cultures are more repressive across the board, and show that the violation of some core rights is actually more common in non-Muslim countries.¹⁵

Finally, the results for our control variables provide support for our theoretical predictions. The coefficient for population was consistently positive and significant across all practices, suggesting that when the population is larger, regimes tend to be more repressive. One possible explanation for this finding, suggested earlier, is that these regimes may be less susceptible to external pressures to avoid repressive practices. An alternative explanation would be that countries with a higher population find it harder to exercise their autonomy and consequently are more likely to resort to repressive measures. The coefficient for GDP per capita was negative and statistically significant (although not for disappearances and political imprisonment), perhaps because higher income countries have at their disposal alternative ways of dealing with dissent and threats, rather than resorting to killings and torture.

Tests for endogeneity

To account for potential endogeneity, we also ran models that we obtained by employing instrumental variable regression techniques. In Table 3, we report the findings from the Arellano–Bond dynamic panel-data models. Results for the Arellano–Bond test show the presence of autocorrelation of the first order and the absence of autocorrelation of the second order, suggesting that the estimator is producing consistent and reliable estimates. Because GMM approaches assume cardinality, the results are easier to interpret than those derived from ordered probit methods. For example, Table 3 shows that an increase of one standard deviation in the average repression score of neighboring countries results in an increase of 0.26 points in the repression score of the focal country.

Table 3.

Arellano–Bond dynamic panel-data estimation, one-step difference generalized method of moments (GMM) analysis of state repression and terrorism, 1981–2005.

	Cumulative Scale^a	Killings	Disappearances	Imprisonment	Torture
Lagged dependent variable	−0.12 (−0.96)	−0.24** (−2.58)	0.00 (0.00)	−0.07 (−0.68)	0.07 (0.75)
Actor-oriented approaches
Terror events (ln)	0.00 (0.46)	0.00 (0.59)	0.01** (2.56)	−0.01* (−1.73)	0.00 (0.13)
Terror events in neighbor states (ln)	0.01 (0.21)	0.05** (2.38)	−0.03 (−1.34)	0.01 (0.43)	−0.01 (−0.56)
Strikes (ln)	0.03 (0.19)	0.05 (0.78)	−0.10* (−1.91)	0.00 (0.06)	0.07 (1.00)
Riots (ln)	−0.11 (−0.63)	0.05 (0.75)	0.02 (0.35)	−0.16* (−1.90)	−0.05 (−0.77)
Revolutions (ln)	0.28** (2.10)	0.11* (1.79)	0.06 (0.94)	0.02 (0.36)	0.06 (1.04)
Demonstrations (ln)	−0.01 (−0.09)	−0.08* (−1.93)	−0.02 (−0.32)	0.07 (1.20)	0.04 (0.79)
Stable regime	−0.22 (−1.27)	−0.14 (−1.66)	0.02 (0.30)	−0.08 (−1.04)	0.00 (0.02)
Civil war	0.41** (2.01)	0.17 (1.64)	0.26** (2.04)	−0.01 (−0.15)	0.00 (−0.01)
International conflicts (ln)	0.01 (0.92)	0.01* (1.83)	0.00 (−0.01)	0.00 (0.27)	0.00 (−0.14)
Socio-cultural approaches
Globalization index	0.03 (1.00)	0.02* (1.93)	0.00 (−0.10)	0.00 (−0.27)	0.02 (1.17)
Human rights treaties ratification	0.14 (0.97)	0.07 (1.15)	0.02 (0.37)	0.06 (0.86)	0.00 (0.02)
Avg. repression in neighbor states	0.27*** (3.44)	0.09*** (2.85)	0.04 (1.27)	0.05* (1.79)	0.09** (2.44)
Democracy	−0.05* (−1.86)	−0.01 (−0.69)	0.00 (0.12)	−0.04*** (−2.70)	0.00 (−0.36)
Controls
Population (ln)	−5.76 (−0.83)	−2.95 (−1.24)	−0.23 (−0.15)	−0.95 (−0.33)	−2.64 (−0.66)
GDP per capita (ln)	−0.89 (−0.84)	−0.13 (−0.37)	−0.29 (−0.75)	−0.22 (−0.53)	−0.37 (−0.78)
Urbanization	−0.06 (−0.28)	−0.09 (−0.65)	−0.20* (−1.66)	0.10 (0.96)	0.18 (1.27)
Hansen test of overid (chi square)	0.02	0.06	0.05	0.07	0.46
Arellano–Bond first-order test	0.00	0.00	0.00	0.00	0.00
Arellano–Bond second-order test	0.11	0.97	0.99	0.24	0.00
Number of instruments	66	66	66	66	66
Number of countries	133	133	133	133	133
Observations	1112	1112	1112	1112	1112

The cumulative scale in the first model is a composite index referring to the violation of physical integrity rights. We reversed original Cingranelli–Richards 9-point scale so that a score of 0 represents low levels of repression while a score of 8 represents high repression levels. The disaggregated measurements of repression (killings, disappearances, imprisonment, and torture) were also reversed, with higher scores now representing higher repression levels. We introduce a control for year in all models to control for linear time trends in repression.

Robust t statistics in brackets. *p < .10; **p < .05; ***p < .01.

The results show that employing instruments changes some of our main findings, in particular the results for terrorist attacks, but also for globalization, population, and GDP per capita. Thus, it appears that once we employ a more conservative correction for endogeneity, terrorism no longer significantly predicts the cumulative scale and only remains a significant predictor of disappearances. These results suggest that endogeneity is a major issue in the association between terrorism and state repression and that the well-established association between the two may in fact be driven mainly by the effects of repression over terrorism rather than the other way around. Furthermore, these models show that while some measures do not withstand a more rigorous test (notably, globalization, population, and GDP per capita, in addition to terrorism), other measures such as revolutions, civil wars, and spatial repression emerge as particularly robust predictors of state repression. Spatial diffusion (as measured by the average repression of neighbor states) is particularly interesting here, as we suspected that reverse causality and endogeneity were especially likely for this predictor. Table 3, however, clarifies that employing instruments does not change our previous finding and repression in neighbor states is clearly a robust predictor of state repression.

Conclusion

This study examined the effects of terrorism and various other factors on state repression of physical integrity rights. The theoretical tension between what Krasner (1999) calls actor-oriented perspectives (or rational choice approaches, as they are often referred to in the literature¹⁶) on the one hand and socio-cultural paradigms on the other hand stood in the center of much of the IR and political sociology literature over the last two decades (Cardenas, 2004; Checkel, 1997; Hafner-Burton and Tsutsui, 2005; Hartigan, 1992; Krasner, 1993; Risse et al., 1999). Scholars who study the relationships between conflicts, insurgency, terrorism, and state policies (Brym and Maoz-Shai, 2009; Mitchell, 2004; Muller and Weede, 1990; Pape, 2005; Ron, 2003; Shor, 2008a, 2008b; Shor and Yonay, 2010, 2011; Yonay and Shor, 2014) also increasingly adopted this analytical framework. While the results of the current study cannot provide a definite resolution to this debate, our findings join a growing body of research (e.g. Brym and Maoz-Shai, 2009; Cole, 2013; Ron, 2000, 2003; Shor, 2008a) that emphasizes the importance of socio-cultural determinants of states’ counterterrorist policies.

Similarly to previous studies (e.g. Dreher et al., 2010; Piazza and Walsh, 2009; Robison, 2010), we found that terror attacks increase state repression of core physical integrity rights, such as killings and imprisonment when adopting a standard ordinal logistic regression panel analysis where covariates are lagged 1 year. Furthermore, our results support the findings presented by Dreher et al. (2010) regarding the disaggregated measures of repression. Terrorism appears to be detrimental for all four of these core human rights. However, when we ran more rigorous models that employ instruments, terrorism was no longer a significant predictor of overall repression and of most of the disaggregated measurements. This suggests that at least when it comes to domestic terrorism, the now well-established association between terrorism and state repression may be largely due to the effects of repression on terrorism rather than the other way around. However, as other variables (e.g. globalization, population, and GDP per capita) also lost their statistical significance when moving to the GMM model, one should be cautious in drawing conclusions and additional examination of the relationship between terrorism and state repression may be in order.

Among the other measurements of traditional actor-oriented IR approaches, two received consistent support in all of our models (including those in which we employed instruments): the number of revolutions and the presence of civil war. In accordance with our third hypothesis, and consistent with the findings of previous studies (Dreher et al., 2010; Piazza and Walsh, 2009; Robison, 2010), both of these were significant predictors of increased repression. In the face of political instability and serious threats to the government’s legitimacy and ability to rule, leaders often resort to repressive measures, in an effort to contain the threat (whether this actually works is a different question).

However, while security threats and political instability play an important role in predicting states’ repressive policies, they cannot provide a comprehensive enough explanation for these policies. Most importantly, actor-oriented explanations fail to consider the socio-institutional and cultural elements that are involved in shaping governmental policies. Indeed, we found consistent support for some socio-institutional approaches. Political, social, and economic globalization were significantly associated with lower levels of physical rights repression, although the effect diminished when we ran models that contained instrumental variables. In addition, spatial diffusion effects were perhaps the single most consistent and robust predictor of state repression in our analysis. Governments’ clearly look to their neighbors for guidance and their respect for core human rights is influenced by the respect of neighboring countries for these rights. It seems therefore that governments often emulate the practices of their neighbors regardless of actual levels of terrorism and other internal or external conflicts and political pressures.

Our results call for an expansion of the way we understand state repression and its antecedents. First, repression is not by any means a reflexive and unavoidable response to terror attacks, as many countries suffering from terrorism choose to deal with the threat with measures that do not infringe on people’s basic rights. Furthermore, it is clear that countries often use real or imagined terrorist threats as an excuse to carry out repressive policies regardless of whether these policies are effective in dealing with the threat. The events over the last few years in the Middle East and North Africa demonstrate this point, as governments in Syria, Egypt, Libya, Yemen, and Bahrain were quick to label the often-nonviolent oppositional protests as a serious terrorist threat and responded with a heavy hand (which often only served to exacerbate the revolt and resulted in further mutual violence).

Footnotes

Appendix

Appendix 1.

List of countries included in the most inclusive analyses.

Afghanistan	Djibouti	Latvia	Serbia and Montenegro
Albania	Dominican Republic	Lesotho	Sierra Leone
Algeria	East Timor	Liberia	Singapore
Angola	Ecuador	Libya	Slovakia
Argentina	Egypt	Lithuania	Slovenia
Armenia	El Salvador	Macedonia	Solomon Islands
Australia	Equatorial Guinea	Madagascar	Somalia
Austria	Eritrea	Malawi	South Africa
Azerbaijan	Estonia	Malaysia	Spain
Bahrain	Ethiopia	Mali	Sri Lanka
Bangladesh	Fiji	Mauritania	Sudan
Belarus	Finland	Mauritius	Swaziland
Belgium	France	Mexico	Sweden
Benin	Gabon	Moldova	Switzerland
Bhutan	Gambia	Mongolia	Syria
Bolivia	Georgia	Morocco	Tajikistan
Botswana	Ghana	Mozambique	Tanzania
Brazil	Greece	Myanmar (Burma)	Thailand
Bulgaria	Guatemala	Namibia	Togo
Burkina Faso	Guinea	Nepal	Trinidad and Tobago
Burundi	Guinea-Bissau	New Zealand	Tunisia
Cambodia	Guyana	Nicaragua	Turkey
Cameroon	Haiti	Niger	Turkmenistan
Canada	Honduras	Nigeria	Uganda
Central African Republic	Hungary	Norway	Ukraine
Chad	India	Oman	United Arab Emirates
Chile	Indonesia	Pakistan	United Kingdom
China	Iran	Panama	United States
Colombia	Iraq	Papua New Guinea	Uruguay
Comoros	Ireland	Paraguay	USSR
Congo (Zaire)	Israel	Peru	Uzbekistan
Congo (Brazzaville)	Italy	Philippines	Venezuela
Costa Rica	Jamaica	Poland	Vietnam
Cote d’Ivoire	Japan	Portugal	Yemen
Croatia	Jordan	Qatar	Yemen, North
Cuba	Kazakhstan	Romania	Yugoslavia
Cyprus	Kenya	Russia	Zambia
Czech Republic	Kuwait	Rwanda	Zimbabwe
Czechoslovakia	Kyrgyzstan	Saudi Arabia
Denmark	Laos	Senegal

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Notes

References

Abouharb

Cingranelli

(2006) The human rights effects of world bank structural adjustment, 1981–2000. International Studies Quarterly 50(2): 233–262.

Achen

(2000) Why lagged dependent variables can suppress the explanatory power of other independent variables. In: Presented at the annual meeting of political methodology, Los Angeles, CA, 20–22 July.

Aktan

Koknar

(2002) Turkey. In: Alexander

(ed.) Combating Terrorism: Strategies of Ten Countries. Ann Arbor, MI: University of Michigan Press, pp. 260–297.

Alexander

(ed.) (2002) Combating Terrorism: Strategies of Ten Countries. Ann Arbor, MI: University of Michigan Press.

Allen

Colley

(2008) Will fear keep them in line? A formal model and empirical testing of the relationship between terrorism and repression. In: Annual meeting of the ISA’s 49th annual convention, San Francisco, CA, 26 March.

Arellano

Bover

(1995) Another look at the instrumental variable estimation of error-components models. Journal of Econometrics 68: 29–51.

Asal

Rethemeyer

(2008) The nature of the beast: Organizational structures and the lethality of terrorist attacks. Journal of Politics 70(2): 437–449.

Banks

(2008) Cross-National Time Series Databanks. Arthur Banks. Available at: http://www.databanksinternational.com/32.html (accessed 3 December 2013).

Barber

(1995) Jihad vs. McWorld: How Globalism and Tribalism are Reshaping the World. New York: Ballantine Books.

10.

Bell

Cingranelli

Murdie

. (2013) Coercion, capacity, and coordination: Predictors of political violence. Conflict Management and Peace Science 30: 240–262.

11.

Bergesen

Lizardo

(2004) International terrorism and the world-system. Sociological Theory 22: 38–52.

12.

Bhoumik

(2005) Democratic responses to terrorism: A comparative study of the United States, Israel, and India. Denver Journal of International Law and Policy 33: 286.

13.

Blankley

(2005) The West’s Last Chance: Will We Win the Clash of Civilizations? Washington, DC: Regnery Publishing.

14.

Boli

Thomas

(1997) World culture in the world polity: A century of international non-governmental organization. American Sociological Review 62: 171–190.

15.

Bonner

(1993) United Kingdom: The United Kingdom response to terrorism. In: Schmid

Crelinsten

(eds) Western Responses to Terrorism. London: Frank Cass, pp. 171–205.

16.

Brachet-Marquez

(1992) Explaining sociopolitical change in Latin America: The case of Mexico. Latin American Research Review 27: 91–122.

17.

Brotóns

Espósito

(2002) Spain. In: Alexander

(ed.) Combating Terrorism: Strategies of Ten Countries. Ann Arbor, MI: University of Michigan Press, pp. 163–186.

18.

Brym

Maoz-Shai

(2009) Israeli state violence during the second intifada: Combining new institutionalist and rational choice approaches. Studies in Conflict & Terrorism 32: 611–626.

19.

Burgoon

(2006) On Welfare and terror: Social Welfare policies and political-economic roots of terrorism. Journal of Conflict Resolution 50(2): 176–203.

20.

Cardenas

(2004) Norm collision: Explaining the effects of international human rights pressure on state behavior. International Studies Review 6: 213–231.

21.

Checkel

(1997) International norms and domestic politics: Bridging the rationalist-constructivist divide. European Journal of International Relations 3: 473–495.

22.

Chomsky

(1988) The Culture of Terrorism. Boston, MA: South End Press.

23.

Chomsky

(1991) International terrorism: Image and reality. In: George

(ed.) Western State Terrorism. New York: Routledge, pp. 12–38.

24.

Cingranelli

Richards

(2010) The Cingranelli-Richards (CIRI) human rights database. Available at: http://www.humanrights.com

25.

Cole

(2013) Does respect for human rights vary across ‘civilizations’? A statistical reexamination. International Journal of Comparative Sociology 54: 345–381.

26.

Davenport

(1995) Multidimensional threat perception and state repression: An inquiry into why states apply negative sanctions. American Journal of Political Science 39: 683–713.

27.

Davenport

(1996) Constitutional promises and repressive realities: A cross-national time series investigation of why political and civil liberties are suppressed. Journal of Politics 58: 627–654.

28.

Davenport

(2007) State repression and political order. Annual Review of Political Science 10: 1–23.

29.

Davenport

Armstrong

(2004) Democracy and the violation of human rights: A statistical analysis from 1976 to 1996. American Journal of Political Science 48: 538–554.

30.

DiMaggio

Powell

(eds) (1991) The New Institutionalism in Organizational Analysis. Chicago, IL: University of Chicago Press.

31.

DiMaggio

Powell

(1983) The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. American Sociological Review 48: 147–160.

32.

Dreher

(2006) Does globalization affect growth? Evidence from a new index of globalization. Applied Economics 38: 1091–1110.

33.

Dreher

Gassebner

Siemers

L-H

(2010) Does terrorism threaten human rights? Evidence from panel data. Journal of Law & Economics 53: 65–93.

34.

Dreher

Gassebner

Siemers

L-H

(2012) Globalization, economic freedom, and human rights. Journal of Conflict Resolution 56: 516–546.

35.

Finkel

Green

Pérez-Liñán

. (2005) Cross-national research on USAID’s democracy and governance programs: Dataset for phase I. Available at: http://www.pitt.edu/~politics/democracy/democracy.html

36.

Fontaine

(2002) Argentina. In: Alexander

(ed.) Combating Terrorism: Strategies of Ten Countries. Ann Arbor, MI: University of Michigan Press, pp. 62–83.

37.

Ganor

(2005) The Counter-Terrorism Puzzle: A Guide to Decision Makers. New Brunswick, NJ: Transaction Publishers.

38.

George

(1991) Introduction. In: George

(ed.) Western State Terrorism. New York: Routledge, pp. 1–11.

39.

Ghosn

Bennett

(2005) Codebook for the Dyadic Militarized Interstate Incident Data, Version 3.10. Available at: http://www.correlatesofwar.org/COW2%20Data/MIDs/Codebook%20for%20Dyadic%20MID%20Data%20v3.10.pdf (accessed 8 September 2012).

40.

Gill

(2004) The School of the Americas: Military Training and Political Violence in the Americas. Durham: Duke University Press.

41.

Goodwin

(2006) A theory of categorical terrorism. Social Forces 84(4): 2027–2046.

42.

Gränzer

(1999) Changing discourse: Transnational advocacy networks in Tunisia and Morocco. In: Risse

Ropp

Sikkink

(eds) The Power of Human Rights: International Norms and Domestic Change. Cambridge: Cambridge University Press, pp. 109–133.

43.

Greenberg

Dratel

(2005) The Torture Papers: The Road to Abu Ghraib. Cambridge: Cambridge University Press.

44.

Gross

(2006) The Struggle of Democracy against Terrorism: Lessons from the United States, the United Kingdom, and Israel. Charlottesville, VA: University of Virginia Press.

45.

Gurr

(1986) The political origins of state violence and terror: A theoretical analysis. In: Stohl

Lopez

(eds) Government Violence and Repression: An Agenda for Research. New York: Greenwood Press, pp. 45–72.

46.

Hafner-Burton

Tsutsui

(2005) Human rights in a globalizing world: The paradox of empty promises. American Journal of Sociology 110: 1373–1411.

47.

Hartigan

(1992) Matching humanitarian norms with cold, hard interests: The making of refugee policies in Mexico and Honduras, 1980–89. International Organization 46: 709–730.

48.

Heinz

(2007) Germany: State responses to terrorist challenges and human rights. In: Brysk

Shafir

(eds) National Insecurity and Human Rights: Democracies Debate Counterterrorism. Berkeley, CA: University of California Press, pp. 157–176.

49.

Heinz

Frühling

(1999) Determinants of Gross Human Rights Violations by State and State-Sponsored Actors in Brazil, Uruguay, Chile, and Argentina, 1960–1990. The Hague: Martinus Nijhoff Publishers.

50.

Henderson

(1991) Conditions affecting the use of political repression. Journal of Conflict Resolution 35: 120–142.

51.

Henderson

(1993) Population pressures and political repression. Social Science Quarterly 74: 322–333.

52.

Hewitt

(ed.) (1984) The Effectiveness of Anti-Terrorist Policies. New York: University Press of America.

53.

Huntington

(1996) The Clash of Civilizations and the Remaking of World Order. New York: Simon & Schuster.

54.

Ikenberry

Kupchan

(1990) Socialization and hegemonic power. International Organization 44: 283–315.

55.

Itabashi

Ogawara

Leheny

(2002) Japan. In: Alexander

(ed.) Combating Terrorism: Strategies of Ten Countries. Ann Arbor, MI: University of Michigan Press, pp. 337–374.

56.

Jimenez

(1993) Spain: The terrorist challenge and the government’s response. In: Schmid

Crelinsten

(eds) Western Responses to Terrorism. London: Frank Cass, pp. 110–130.

57.

Keck

Sikkink

(1998) Activists beyond Borders: Advocacy Networks in International Politics. Ithaca, NY: Cornell University Press.

58.

Kramer

(2005) The perils of counterinsurgency: Russia’s war in Chechnya. International Security 29(3): 5–63.

59.

Krasner

(1993) Sovereignty, regimes, and human rights. In: Rittberger

(ed.) Regime Theory and International Relations. Oxford: Oxford University Press, pp. 139–167.

60.

Krasner

(1999) Sovereignty: Organized Hypocrisy. Princeton, NJ: Princeton University Press.

61.

Kuper

(1981) Genocide. London: Penguin Books.

62.

Lacquer

(1999) The New Terrorism: Fanaticism and the Arms of Mass Destruction. New York: Oxford University Press.

63.

Lacquer

Alexander

(eds) (1987) The Terrorism Reader: A Historical Anthology. New York: New American Library.

64.

LaFree

Dugan

(2008a) Global terrorism database, 1970–1997 (Computer file). ICPSR04586-v1. College Park, MD: University of Maryland. Available at: http://www.start.umd.edu/

65.

LaFree

Dugan

(2008b) Global terrorism database, 1998–2004 (Computer file). ICPSR04586-v2. College Park, MD: University of Maryland. Available at: http://www.start.umd.edu/

66.

Lai

(2007) ‘Draining the swamp’: An empirical examination of the production of international terrorism, 1968–1998. Conflict Management and Peace Science 24(4): 297–310.

67.

Lustick

(2006) Trapped in the War on Terror. Philadelphia, PA: University of Pennsylvania Press.

68.

Lutz

Sikkink

(2000) International human rights law and practice in Latin America. International Organization 54: 633–659.

69.

Marshall

(2014) Polity IV Project: Political Regime Characteristics and Transitions, 1800–2013. Available at: http://www.systemicpeace.org/polity/polity4.htm

70.

Marti

Domingo

Ibarra

(2007) Democracy, civil liberties, and counterterrorist measures in Spain. In: Brysk Shafir

(eds) National Insecurity and Human Rights: Democracies Debate Counterterrorism. Berkeley, CA: University of California Press, pp. 118–137.

71.

Marwah

(2002) India. In: Alexander

(ed.) Combating Terrorism: Strategies of Ten Countries. Ann Arbor, MI: University of Michigan Press, pp. 301–336.

72.

Meyer

Rowan

(1977) Institutionalized organizations: Formal structures as myth and ceremony. American Journal of Sociology 83: 340–363.

73.

Meyer

Scott

(1992) Organizational Environments: Ritual and Rationality. Beverly Hills, CA: SAGE.

74.

Meyer

Boli

Thomas

. (1997) World Society and the Nation-state. American Journal of Sociology 103: 144–181.

75.

Mitchell

(2004) Agents of Atrocity: Leaders, Followers, and the Violation of Human Rights in Civil War. New York: Palgrave Macmillan.

76.

Mitchell

McCormick

(1988) Economic and political explanations of human rights violations. World Politics 40: 476–498.

77.

Moravcsik

(1995) Explaining international human rights regimes: Liberal theory and Western Europe. European Journal of International Relations 1: 157–189.

78.

Moravcsik

(1997) Taking preferences seriously: A liberal theory of international politics. International Organization 51: 513–553.

79.

Moravcsik

(2000) The origins of human rights regimes: Democratic delegation in postwar Europe. International Organization 54: 217–252.

80.

Muller

Weede

(1990) Cross-national variation in political violence: A rational action approach. Journal of Conflict Resolution 34(4): 634–651.

81.

Murphy

(2006) United States Practice in International Law. Cambridge: Cambridge University Press.

82.

Okudzeto

(2007) Stamping Out Rights: The Impact of Anti-Terrorism Laws on Policing. New Delhi, India: Commonwealth Human Rights Initiative (CHRI). Available at: www.humanrightsinitiative.org

83.

Pape

(2005) Dying to Win: The Strategic Logic of Suicide Terrorism. New York: Random House.

84.

Piazza

(2006) Rooted in poverty? Terrorism, poor economic development and social cleavages. Terrorism and Political Violence 18: 159–177.

85.

Piazza

Walsh

(2009) Transnational terror and human rights. International Studies Quarterly 53: 125–148.

86.

Pion-Berlin

Lopez

(1991) Of victims and executions: Argentine state terror, 1975–1979. International Studies Quarterly 35: 63–86.

87.

Poe

Tate

(1994) Repression of human rights to personal integrity in the 1980s: A global analysis. American Political Science Review 88: 853–872.

88.

Poe

Tate

Keith

(1999) Repression of the human right to personal integrity revisited: A global cross-national study covering the years 1976–1993. International Studies Quarterly 43: 291–313.

89.

Pridham

(1981) Terrorism and the state in West Germany during the 1970s: A threat to stability of a case of political over-reaction? In: Lodge

(ed.) Terrorism: A Challenge to the State. New York: St. Martin’s Press, pp. 11–56.

90.

Risse

Ropp

Sikkink

(eds) (1999) The Power of Human Rights: International Norms and Domestic Change. Cambridge: Cambridge University Press.

91.

Robinson

(1995) Strengthening civil society in Africa: The role of foreign political aid. IDS Bulletin: Institute of Development Studies 26: 70–80.

92.

Robison

(2006) Civil and state terrorism: Terrorism’s effects on states’ respect for human rights. In: Annual meeting of the American Sociological Association, Montreal, QC, Canada, 11 August.

93.

Robison

(2007) The challenges of political terrorism: A cross-national analysis of the downward spiral of terrorist violence and socio-political crisis. In: Sociology. Columbus, OH: Ohio State University, 230 pp.

94.

Robison

(2010) Terror’s true nightmare? Reevaluating the consequences of terrorism on democratic governance. Terrorism and Political Violence 22: 62–86.

95.

Robison

Crenshaw

Craig

(2006) Ideologies of violence: The social origins of Islamist and Leftist transnational terrorism. Social Forces 84(4): 2009–2026.

96.

Roniger

Sznajder

(1999) The Legacy of Human-Rights Violations in the Southern Cone: Argentina, Chile, and Uruguay. Oxford: Oxford University Press.

97.

Ron

(2000) Savage restraint: Israel, Palestine and the dialectics of legal repression. Social Problems 47: 445–472.

98.

Ron

(2003) Frontiers and Ghettos: State Violence in Serbia and Israel. Berkeley, CA: University of California Press.

99.

Roodman

(2006) How to do xtbond2: An introduction to ‘difference’ and ‘system’ GMM in Stata. Working Paper no. 103, December. Washington, DC: Center for Global Development. Available at: http://www.cgdev.org/files/11619_file_HowtoDoxtabond6_12_1_06.pdf

100.

Sanchez-Cuenca

de la Calle

(2009) Domestic terrorism: The hidden side of political violence. Annual Review of Political Science 12: 31–49.

101.

Scheppele

(2005) ‘We forgot about the ditches’: Russian constitutional impatience and the challenge of terrorism. Drake Law Review 53: 963–1027.

102.

Shandra

Shor

London

(2008) World polity, unequal ecological exchange, and organic water pollution: A cross-national analysis. Human Ecology Review 16: 52–62.

103.

Shandra

Shor

London

(2009) Debt, structural adjustment, and organic water pollution: A cross-national analysis. Organization & Environment 21: 38–55.

104.

Shor

(2008a) Conflict, terrorism, and the socialization of human rights norms: The spiral model revisited. Social Problems 55: 117–138.

105.

Shor

(2008b) Contested masculinities: The new Jew and the construction of black and Palestinian athletes in Israeli media. Journal of Sport & Social Issues 32: 255–277.

106.

Shor

(2008c) Utilizing rights and wrongs: Right-wing, the ‘right’ language, and human rights in the Gaza disengagement. Sociological Perspectives 51: 803–826.

107.

Shor

(2010) Terrorism and human rights in the new Millennium: In search for equilibrium (a review article). Qualitative Sociology 33: 205–210.

108.

Shor

(2011) Constructing a global counterterrorist legislation database: Dilemmas, procedures, and preliminary analyses. Journal of Terrorism Research 2: 249–277.

109.

Shor

Yonay

(2010) Sport, national identity, and media discourse over foreign athletes in Israel. Nationalism & Ethnic Politics 16: 483–503.

110.

Shor

Yonay

(2011) ‘Play and shut up’: The silencing of Palestinian athletes in Israeli media. Ethnic and Racial Studies 34: 229–247.

111.

Sikkink

(2004) Mixed Signals: U.S. Human Rights Policy and Latin America. Ithaca, NY: Cornell University Press.

112.

Sikkink

Booth Walling

(2006) Argentina’s contribution to global trends in transitional justice. In: Roht-Arriaza

Mariezcurrena

(eds) Transitional Justice in the Twenty-First Century: Beyond Truth versus Justice. Cambridge: Cambridge University Press, pp. 301–324.

113.

Smith

Pagnucco

Lopez

(1998) Globalizing human rights: The work of transnational human rights NGOs in the 1990s. Human Rights Quarterly 20: 379–412.

114.

Sobieck

(1994) Democratic responses to international terrorism in Germany. In: Charters

(ed.) The Deadly Sin of Terrorism: It’s effect on Democracy and Civil Liberty in Six Countries. Westport, CT: Greenwood Press, pp. 43–71.

115.

Stepan

Robertson

(2003) An ‘Arab’ more than ‘Muslim’ democracy gap. Journal of Democracy 14: 30–44.

116.

The Peace Research Institute Oslo (PRIO) (2009) Center for Study of Civil War Data. The Peace Research Institute Oslo (PRIO).

117.

Tsutsui

(2004) Global civil society and ethnic social movements in the contemporary world. Sociological Forum 19: 63–88.

118.

Van de Voorde

(2005) Sri Lankan terrorism: Assessing and responding to the threat of the iberation Tigers of Tamil Eelam (LTTE). Police Practice and Research 6: 181–199.

119.

Vreeland

(2008) Political Institutions and Human Rights: Why dictatorships enter into the United Nations convention against torture. International Organization 62: 65–101.

120.

Walsh

Piazza

(2010) Why respecting physical integrity rights reduces terrorism. Comparative Political Studies 43: 551–577.

121.

Waltz

(1979) Theory of International Politics. Reading, MA: Addison-Wesley.

122.

Wendt

(1999) Social Theory of International Politics. Cambridge: Cambridge University Press.

123.

Whittaker

(2001) The Terrorism Reader. New York: Routledge.

124.

Wikipedia (2009) List of Members of the Commonwealth of Nations. Wikipedia.

125.

Wilkinson

(2001) Terrorism versus Democracy: The Liberal State Response. Portland, OR: Frank Cass.

126.

World Bank (2007) World Development Indicators. Washington, DC: Compact Disk.

127.

World Resources Institute (2006) EarthTrends: Environmental Information. Washington, DC: World Resources Institute.

128.

Yonay

Shor

(2014) Ethnic coexistence in deeply divided societies: The case of Arab athletes in the Hebrew media. Sociological Quarterly 55: 396–420.