ArticlePDF Available

Middle East and North Africa: Terrorism and Conflicts


Abstract and Figures

During 2002–2018, the Middle East and North Africa (MENA) accounted globally for 36.1 per cent of terrorist incidents, 49.3 per cent of terrorist‐induced casualties, and 21.4 per cent of conflict deaths. One focus here is to investigate how MENA's terrorist attacks and conflicts compare with those in the world's other six regions during selected periods, drawn from 1970–2018. There is a well‐defined shift of terrorism from Latin America, Europe and Central Asia to MENA, South Asia, and sub‐Saharan Africa after 1989. A second focus is to employ panel regressions to contrast the drivers of global terrorism with those of MENA terrorism. Democracy and civil conflicts are main drivers of MENA terrorism, followed by population. Regional peacekeeping can have an ameliorating effect on terrorism by limiting conflict. The Arab Spring and associated regime changes are shown to have ushered in a wave of terrorism in MENA. Policy recommendations conclude the study.
Content may be subject to copyright.
Middle East and North Africa: Terrorism and
Wukki Kim
Korea Military Academy
Todd Sandler
University of Texas at Dallas
During 20022018, the Middle East and North Africa (MENA) accounted globally for 36.1 per cent of terrorist incidents, 49.3
per cent of terrorist-induced casualties, and 21.4 per cent of conict deaths. One focus here is to investigate how MENA's ter-
rorist attacks and conicts compare with those in the world's other six regions during selected periods, drawn from 1970
2018. There is a well-dened shift of terrorism from Latin America, Europe and Central Asia to MENA, South Asia, and sub-
Saharan Africa after 1989. A second focus is to employ panel regressions to contrast the drivers of global terrorism with those
of MENA terrorism. Democracy and civil conicts are main drivers of MENA terrorism, followed by population. Regional peace-
keeping can have an ameliorating effect on terrorism by limiting conict. The Arab Spring and associated regime changes are
shown to have ushered in a wave of terrorism in MENA. Policy recommendations conclude the study.
Policy Implications
During 19702018, terrorist attacks have greatly shifted from Latin America and the Caribbean (LAC) and Europe and Cen-
tral Asia (ECA) to the Middle East and North Africa (MENA), South Asia (SAS), and sub-Saharan Africa (SSA), so that
enhanced counterterrorism vigilance and resources are particularly needed in SAS, SSA, and MENA.
Two main determinants of domestic and transnational terrorism in MENA have been democracy and the post-Arab Spring
era. In MENA, greater democracy has not been a panacea for terrorism, contrary to the beliefs of the George W. Bush
administration. The international community must build up the necessary institutional infrastructure to support democracy
in the region.
MENA is plagued by three major civil wars and other insurgencies that foment regional instability, so that international
peacekeeping efforts are needed to curtail these conicts. Given the huge importance of the number of civil conicts in
causing terrorism in the region, such peacekeeping also serves as a counterterrorism measure.
Neither reduced poverty (low GDP per capita) nor defense spending ameliorates terrorism in MENA so that donor and
local countriespolicy makers must look beyond foreign aid or increased defense spending to address terrorism.
Domestic terrorism poses a far-greater and growing threat than transnational terrorism in MENA and worldwide. Efforts to
ght domestic terrorism require donor-country partnerships with terrorism-plagued MENA countries. Despite their dif-
culty, such partnerships must be forged even though domestic terrorism may not pose an immediate threat to the donors
interests at home or on foreign soil.
Since 1989, the Middle East and North Africa (MENA) has
been plagued by growing terrorist attacks and conicts,
fueling signicant increases in the regions military expendi-
ture (ME) (Sandler and George, 2016; SIPRI, 2019; Stockholm
International Peace Research Institute (SIPRI) 2017, 2019).
For instance, MENAs global share of terrorist attacks rose
from 9.8 per cent during 19701989 to 36.1 per cent during
20022018 see Section 2.
In this latter period, MENA
accounted for 21.4 per cent of global conict deaths. The
political violence and unrest in MENA had negative conse-
quences for other regions stemming from spillover terrorism
abroad, foreign direct investment (FDI) losses, disrupted oil
exports, reduced economic growth, and large refugee ows.
Relatively few studies focus on MENA in terms of its terror-
ism and conict. This relative absence of up-to-date terror-
ism and conict analyses for MENA is quite surprising
because this region has been the birthplace for many noto-
rious terrorist groups (e.g., the Popular Front for the Libera-
tion of Palestine [PFLP], Black September, Fatah, Abu Nidal
Organization, Hezbollah, al-Qaida in Iraq, al-Nusra, al-Qaida
in the Arabian Peninsula [AQAP], and Islamic State [IS]) and
has been embroiled in conicts over the last decades.
In a pathbreaking and related study, Piazza (2007) focuses
on the determinants of terrorism in MENA during 1972
2003 and, like the current study, warns that democracy may
worsen terrorism in the region. There are, however, many
Global Policy (2020) doi: 10.1111/1758-5899.12829 ©2020 The Authors. Global Policy published by Durham University and John Wiley & Sons Ltd.
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and
distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modications or adaptations are made.
Global Policy
Research Article
noteworthy distinctions between his article and ours. First,
his sample period ended in 2003 well before the start of the
Arab Spring in December 2010. Second, for MENA, Piazza
(2007) examines the determinants of transnational and total
terrorism, while we investigate the determinants of domestic
and transnational terrorism. Third, unlike the earlier study,
we contrast the drivers of terrorism for a global sample with
those for MENA. Fourth, we can highlight these drivers for
20022018 for current relevance. Fifth, we include some
novel controls as drivers including the post-Arab Spring era,
the number of civil conicts, an economic freedom index,
defense spending burdens, and the number of resident reli-
gious fundamentalist terrorist groups. The number of civil
conicts is an especially important driver owing to the link-
age between terrorism and these conicts.
From a conict vantage, consider the following: In recent
years, nine of the 20 MENA countries deployed combat
forces on their own territory (SIPRI, 2017). In Algeria, Egypt,
Iraq, Lebanon, and Tunisia, internal conicts involved IS,
while in Iran, internal conict concerned Kurdish armed
groups (SIPRI, 2017). Currently, Libya, Syria, and Yemen
endure civil wars with international interventions from the
Middle East, Europe, Russia, and the United States (SIPRI,
2019). Ten MENA countries Bahrain, Egypt, Iran, Israel, Jor-
dan, Kuwait, Morocco, Qatar, Saudi Arabia, and the United
Arab Emirates (UAE) participated in one or more of these
civil wars during recent years. The tensions between Saudi
Arabia and Iran, founded on religious and ideological differ-
ences, sparked conicts throughout the region.
Many MENA countries suffered from a so-called resource
cursethat often, but not always, results in low economic
growth, authoritarian rule (e.g., Iran, Saudi Arabia, Syria, and
UAE), inationary pressures, or large military burdens.
burdens, as measured by the ratio of military expenditure
(ME) to gross domestic product (GDP), stem from a per-
ceived need to protect resource wealth. MENA countries
import their defense systems from varied suppliers with the
United States, Russia, France, Germany, the United Kingdom,
and China being the major exporters of arms to the region
during 20142018 (SIPRI, 2019). Major MENA arms importers
include Saudi Arabia, Egypt, Algeria, UAE, and Iraq, in
decreasing order.
Because of its regional conicts, political violence, and
vast oil wealth, MENA plays a key role in the prospects for
global terrorism, world tranquility, and economic prosperity.
An essential purpose of the current paper is to analyze the
nature of the terrorist threat originating in the region since
1970 relative to the other six World Banks (2019) desig-
nated regions South Asia (SAS), Latin America and the Car-
ibbean (LAC), Europe and Central Asia (ECA), East Asia and
Pacic (EAP), sub-Saharan Africa (SSA), and North America
(NA) in terms of their changing shares of terrorist inci-
dents and casualties (i.e., deaths and injuries). Unlike other
regions, North America includes just two countries Canada
and the United States. We employ panel regressions to
identify the drivers of MENA terrorism and distinguish them
from those associated with global terrorism. In so doing, key
similarities and differences of these drivers at the global and
MENA levels are identied. Another essential purpose is to
highlight changes in conicts by regions over time. Like ter-
rorism, conicts shifted their primary regional focus to SAS,
SSA, and MENA during the post-9/11 era. Because terrorism
and civil conict may occur in tandem, terrorisms regional
shifts may mimic conictsregional movement (Findley and
Young, 2012). Civil conicts may induce terror attacks, but
these attacks seldom cause civil conicts. On average, terror
attacks involve a small number of casualties unlike domestic
There are several ndings to highlight. First, the large
geographic displacement in the venue of terrorism from
Latin America and Europe to the Middle East, Africa, and
South Asia has signicant implications for the practice of
counterterrorism in terms of intelligence, international coop-
eration, conict aid, and antiterrorism resource preposition-
ing. Second, there are important differences between the
drivers of global and MENA terrorism. The latter is especially
bolstered by democracy, civil conicts, and the post-Arab
Spring era. Third, the recent rising share of domestic terror-
ism is less likely to motivate strong Western countries to
provide counterterrorism assistance. Fourth, the changing
location of internal conicts affects refugee ows, resource
supplies, and peacekeeping missions. Fifth, by gaining a per-
spective on MENA terrorism drivers, the global community is
better able to decide how best to ameliorate instabilities in
the region. Given the importance of conict as a driver of
terrorism, reduced arms exports to MENA may foster great
tranquility and reduced terrorism.
Terrorism is the premeditated use or threat to use violence
by individuals or subnational groups to obtain a political or
social objective through the intimidation of a large audience
beyond that of the immediate victims(Enders and Sandler,
2012, p. 4). This denition emphasizes that subnational enti-
ties, and not the state, are the terrorists, thereby ruling out
state terror. In the above denition, the political motive of
perpetrators is an essential ingredient. Without this motive,
a tactic such as kidnapping for ransom is a criminal act of
extortion rather than a terrorist incident to further or fund a
political agenda. The same is true for other modes of attack
such as bombings, assassinations, and armed attacks.
Another crucial component of terrorism is audience intimi-
dation, intended to mobilize public pressure on the govern-
ment to concede to terroristsdemands for political change
to end the terror campaign. The denition here concurs
with those utilized by the two major terrorist event data
sets namely, International Terrorism: Attributes of Terrorist
Events (ITERATE) (Mickolus et al., 2019) and Global Terrorism
Database (GTD) (National Consortium for the Study of Ter-
rorism and Responses to Terrorism (START) 2019).
Terrorist attacks can be decomposed into two main cate-
gories. Domestic terrorism involves victims and perpetrators
whose citizenship is exclusively from the venue country
hosting the incident. Moreover, domestic terrorist attacks
have consequences for just the host or venue country, its
©2020 The Authors. Global Policy published by Durham University and John Wiley & Sons Ltd. Global Policy (2020)
Wukki Kim and Todd Sandler
institutions, citizens, and property. For instance, the assassi-
nation of a Lebanese ofcial by a Lebanese terrorist to inu-
ence local politics is a domestic terrorist attack. Most
countries can be self-reliant in addressing domestic terrorist
incidents if their governments possess enough resources.
Unless there is an eventual spillover of terrorism to neigh-
boring or other countries, domestic antiterrorist policies do
not have to involve foreign governments. The key exception
is when domestic terrorism presents a clear danger to the
host governments stability, which, in turn, affects trade
ows, resource supplies, and other interests abroad. In such
instances, foreign donor countries may provide a terrorism-
plagued country with conict aid to address domestic ter-
rorism (Bandyopadhyay et al., 2011).
Transnational terrorism constitutes the second kind of ter-
rorism when an attack in one country involves perpetrators,
victims, institutions, governments, or citizens from another
country. If a terror attack includes perpetrators from the
venue country and one or more victims from another coun-
try, then the attack is transnational. Transnational incidents
may also involve an attack that starts in one country and
concludes in another. The coordinated attacks in Paris on 13
November 2015, involving suicide bombers outside a foot-
ball match in Saint-Denis, armed attacks at the Bataclan
Theater, and shootings at cafes and restaurants, are transna-
tional terrorist attacks. The perpetrators were from terrorist
cells in Belgium, who crossed a border before and after the
attacks. At the Bataclan Theater, victims came from 17 dif-
ferent countries. Moreover, IS, a foreign terrorist group,
organized the attack, which had been planned outside of
France. In response to the attacks, the French later carried
out bombing raids on IS targets in Syria. Thus, the Paris
attacks had many international linkages. With transnational
terrorism, countriescounterterrorism policies become inter-
dependent. Efforts by the United States and European coun-
tries to secure their borders and ports of entry following 9/
11 or the 11 March 2004 Madrid commuter train bombings
may merely displace future attacks to US and European
interests in MENA, where borders may be more porous
(Enders and Sandler, 1993, 2006, 2012).
For our comparative investigation of terrorism in MENA
and other regions, we focus on three key time periods. In
regards to 19701989, the initial date closely corresponds to
the start of the modern era of transnational terrorism in
1968 when terrorist groups (e.g., PFLP and Black September)
staged major terrorist attacks (e.g., transnational skyjackings
or the 1972 abduction of Israeli athletes at the Munich
Olympics) in foreign cities to publicize their cause to a glo-
bal community (Hoffman, 2006). Terrorist groupstransna-
tional strategy was fostered by advances in communication,
especially satellite television transmissions in the 1960s. This
19701989 time period not only includes the dominance of
the leftist terrorists, but also the bloody era of state-sponsor-
ship in the 1980s (Hoffman, 2006). Both the PFLP and the
Abu Nidal Organization engaged in prominent state-spon-
sored terrorist incidents. The second noteworthy period runs
from 1990 to 2001 and encompasses the rising prominence
of the religious fundamentalist terrorists and the decline of
the leftist terrorists (Enders and Sandler, 2000, 2012; Hoff-
man, 2006; Rapoport, 2004). This period concludes with the
four skyjackings on 9/11, leaving almost 3,000 dead and
constituting the largest terrorist incident in the modern era.
Finally, our third focus period is 20022018, which witnesses
the dominance of religious fundamentalist terrorists, the
continued presence of nationalist terrorists, enhanced post-
9/11 border security, and the decline in transnational terror-
ism (Gaibulloev and Sandler, 2019; Hou et al., 2020). When
we present our panel regressions, we focus on just two peri-
ods 19702018 and 20022018 to ensure adequate
observations to identify signicant drivers of domestic and
transnational terrorism at the global and MENA levels.
Anal preliminary involves the component countries for
each of the seven World Bank (2019) regional classications.
These are given in Table 1, along with their regional abbre-
viation for easy reference.
Changing terrorism threat in MENA relative to
other regions
To offer some context to our subsequent panel regressions
of the determinants of terrorism, we present an overview of
terrorism for three essential time periods. This overview
involves the regional distribution of terrorist incidents and
casualties, along with the changing composition of total ter-
rorism into its domestic and transnational components.
Then we turn to the breakdown of conicts, a key driver of
terrorism, for the seven regions of the world and for the
countries of MENA.
Although Americans have been a prime target of terror-
ism since 1970, most attacks against US interests take place
on foreign soil in terms of transnational terrorism (Gaibul-
loev and Sandler, 2019). Figure 1 displays the shifting
regional distribution of terrorist incidents domestic and
transnational combined for the three highlighted periods.
In moving forward in time, there are some unmistakable
patterns in regional venues, revealed by START (2019) GTD
event data.
For simplicity, the percentages, and not the
count, of total terrorist attacks for each region in the three
periods are shown. Over time, MENA, SAS, and SSA
accounted for an increasing percentage of attacks, while
LAC, ECA, and NA accounted for a decreasing percentage
of attacks. The huge percentage decreases in LAC and ECA
are particularly noteworthy in terms of the allocation of
counterterrorism resources since 2002. A somewhat mixed
pattern applied to EAP. By 20022018, just over 70 per
cent of terrorism occurred in MENA and SAS combined,
with SSA, EAP, and ECA experiencing 13.4 per cent, 9.2 per
cent, and 4.6 per cent of incidents, respectively. Just over 2
per cent and less than 1 per cent of incidents took place
in LAC and NA, respectively, after 2001. The last percent
indicates that the United States must protect its interests
against terrorism abroad, especially after its security
upgrades following 9/11.
The changing geographical distribution of terrorism was
driven by several considerations. The shift of terrorist events
to MENA, SAS, and SSA was due to the rise of religious
Global Policy (2020) ©2020 The Authors. Global Policy published by Durham University and John Wiley & Sons Ltd.
Middle East and North Africa 3
fundamentalist terrorism in Afghanistan, Egypt, Iraq, Libya,
Mali, Nigeria, Pakistan, Somalia, Sudan, Syria, Yemen, and
elsewhere. In the case of MENA, regional civil wars and the
Arab Spring instability reinforced the rise in terrorism. Post-
9/11 enhanced security in NA and parts of ECA was also
responsible for some of those regional redistributions. Since
19701989, the ECA terror decline was also tied to the fall
of communism and the reduction in state-sponsorship ter-
rorism, especially in Europe (Enders and Sandler, 1999; Hoff-
man, 2006). The reduction in leftist and nationalist terrorism
in South America drove the falling percentage of attacks in
In Figure 2, we decompose terrorism carnage in percent-
age terms for attacks taking place in the seven regions for
the three selected time periods. Percentages of regional
casualties are calculated from START (2019) GTD terrorist
event data. Understandably, the intertemporal patterns in
Figure 2 are like those in Figure 1. As time moves forward,
the regional percentages of casualties increased in MENA
and SSA, while these percentages decreased in LAC and
ECA. For SAS, the earlier pattern from Figure 1 is broken
somewhat in Figure 2, where the middle period displayed a
slightly greater regional percentage of casualties than for
20022018. This slight aberration comes from our use of
percentages; SAS casualties increased fourfold from just
under 50,000 during 19902001 to over 205,000 during
20022018. However, the huge increase of casualties in
MENA to 337,432 is responsible for the slight fall in SASs
relative percentage of the global total during 20022018.
The mixed pattern for NA is attributed to the large carnage
Table 1. List of countries in dataset based on the World Bank regional classication
East Asia and Pacic (EAP)
Australia Fiji Lao PDR New Zealand Thailand
Brunei Darussalam Indonesia Malaysia Papua New Guinea Timor-Leste
Cambodia Japan Mongolia Philippines Vietnam
China Korea, Rep. Myanmar Singapore
Europe and Central Asia (ECA)
Albania Czech Republic Ireland Netherlands Sweden
Armenia Denmark Italy Norway Switzerland
Austria Estonia Kazakhstan Poland Tajikistan
Azerbaijan Finland Kyrgyz Republic Portugal Turkey
Belarus France Latvia Romania Turkmenistan
Belgium Georgia Lithuania Russia Ukraine
Bosnia and Herzegovina Germany Luxembourg Serbia United Kingdom
Bulgaria Greece Macedonia Slovak Republic Uzbekistan
Croatia Hungary Moldova Slovenia
Cyprus Iceland Montenegro Spain
Latin America and Caribbean (LAC)
Argentina Colombia Guatemala Mexico Trinidad and Tobago
Belize Cuba Guyana Nicaragua Uruguay
Bolivia Dominican Republic Haiti Panama Venezuela, RB
Brazil Ecuador Honduras Paraguay
Chile El Salvador Jamaica Peru
Middle East and North Africa (MENA)
Algeria Iran Kuwait Morocco Syria
Bahrain Iraq Lebanon Oman Tunisia
Djibouti Israel Libya Qatar United Arab Emirates
Egypt, Arab Rep. Jordan Malta Saudi Arabia Yemen, Rep.
North America (NA)
Canada United States
South Asia (SAS)
Afghanistan Bhutan Nepal Pakistan Sri Lanka
Bangladesh India
Sub-Saharan Africa (SSA)
Angola Congo, Dem. Rep. Guinea Mauritius Somalia
Benin Congo, Rep. Guinea-Bissau Mozambique South Africa
Botswana Cote dIvoire Kenya Namibia South Sudan
Burkina Faso Equatorial Guinea Lesotho Niger Sudan
Burundi Eritrea Liberia Nigeria Tanzania
Cabo Verde Ethiopia Madagascar Rwanda Togo
Cameroon Gabon Malawi Senegal Uganda
Central African Rep. Gambia, The Mali Seychelles Zambia
Chad Ghana Mauritania Sierra Leone Zimbabwe
©2020 The Authors. Global Policy published by Durham University and John Wiley & Sons Ltd. Global Policy (2020)
Wukki Kim and Todd Sandler
of 9/11, that is, the 7,423 casualties during 19902001 was
mostly due to the deaths and injuries from 9/11 in the Uni-
ted States. The pattern for EAP is like that of Figure 1. Gen-
erally, the drivers behind Figure 2 as those of Figure 1.
During 20022018, the regions of most concern are MENA,
SAS, and SSA for which MENA accounts for almost half of all
terrorist casualties, thus justifying our focus on this region.
Almost 80 per cent of all terrorist casualties are in MENA
and SAS, thereby providing a clear roadmap for where
counterterrorism resources are most needed today.
Using the method devised by Enders et al. (2011), we
decompose GTD terrorist events into domestic and transna-
tional attacks to offer yet another perspective on terrorism
in MENA and the other six regions. As shown in Figure 3,
the share of domestic terrorist incidents
for all seven
regions increases from 19902001 to 20022018 due, in
part, to post-9/11 security augmentation. MENAs share of
domestic terrorist attacks goes from 69 per cent in 1970
1989 to 95 per cent in 20022018 so that only one in
twenty attacks involves foreign interests. As such, non-MENA
countries may be unmotivated to do much about MENAs
recent terrorism given its apparent insular nature, which
may be short-sighted if domestic terrorism eventually
morphs into transnational terrorism as shown by Enders
et al. (2011). By 20022018, domestic shares of terrorism in
the other six regions were as follows: SAS, 95 per cent; LAC,
93 per cent; ECA, 68 per cent; EAP, 96 per cent; SSA, 88 per
cent; and NA, 96 per cent. Those shares underscore that
domestic terrorism is the overriding concern in all regions
except Europe and Central Asia.
That concern is bolstered
because domestic terrorist incidents display a greater pro-
portion of attacks with casualties than transnational inci-
dents (Gaibulloev and Sandler, 2019; Gaibulloev et al., 2012).
Although the international media emphasize the threat of
transnational terrorism, this focus is misplaced in recent
years as domestic terrorism accounted for the lions share of
attacks. The predominance of domestic terrorism means that
donor countries are less motivated to curb terrorism in aid-
recipient countries because their resident terrorist groups do
not export many incidents or harm much foreign interests
on the recipients soil except in ECA.
Within-region conicts are apt to fuel terrorist attacks,
particularly domestic terrorism; however, domestic terrorism
does not generally fuel domestic conict. To motivate the
importance of conict as an independent variable in our
subsequent regressions, we take a closer look at the regio-
nal distribution of conict deaths for just two intervals
19892001 and 20022018 based on Uppsala Conict
Data Program (UCDP) data (Sundberg and Melander, 2013).
We remove deaths from Rwanda, which is a huge outlier
given its human toll of over half a million. In Figure 4, the
regional shares of deaths rose in SAS, MENA, and LAC, while
these shares fell in SSA and ECA from the rst to second
period. The major SSA conict in the initial period was the
war between Ethiopia and Eritrea with 87,692 and 79,452
deaths in 19891991 and 19992000, respectively.
in Sudan, Democratic Republic of Congo, Angola, Sierra
Leone, and Somalia also contributed SSA conict deaths in
Figure 1. Regional distribution of terrorist incidents for three time periods.
Global Policy (2020) ©2020 The Authors. Global Policy published by Durham University and John Wiley & Sons Ltd.
Middle East and North Africa 5
the pre-2002 period. Although conict deaths fell in SSA
after 2001, this region still accounted for 29.1 per cent of
the global total in 20022018 when conicts in Sudan, Nige-
ria, Democratic Republic of the Congo, and Somalia were
the main sources of the violence. SAS and MENA are tied to
the largest percentage growth in conict deaths between
the two periods. For SAS, conict deaths during the pre-
2002 period came from the Afghan governments
Figure 2. The regional shares of casualties by terrorist incidents for three time periods.
1990-2001 2002-2018
Figure 3. Share of domestic terrorist incidents for three time periods.
1970-1989 1990-2001 2002-2018
©2020 The Authors. Global Policy published by Durham University and John Wiley & Sons Ltd. Global Policy (2020)
Wukki Kim and Todd Sandler
engagement with the United Islamic Front for Salvation of
Afghanistan, al-Qaida, and the Taliban. Conict deaths in
SAS were also from the Tamil Tigers insurgency in Sri Lanka
and the Kashmir insurgency in India in the pre-2002 period.
In the ensuing period, SASs conict deaths arose from the
war and subsequent insurgency in Afghanistan, the insur-
gency in Sri Lanka, and Taliban-related conict in Pakistan.
MENA conict deaths came from the Gulf War (19901991),
Iraqs post-war insurgency, and the Algerian civil war during
the initial period, while MENA conict deaths stemmed from
the US-led Iraq invasion (2003), regional civil wars (Libya,
Syria, and Yemen), and IS attacks in the subsequent period.
During the most recent period, 87 per cent of conict
deaths were contributed by SAS, SSA, and MENA. By con-
sulting the terrorism-related casualties in Figure 2, we see
that those three regions also accounted for the most human
toll. This suggests that there is a link between terrorism-
and conict-induced casualties; however, the link is by no
means perfect. The connection is greatly blurred when war,
but not insurgency, was the main factor behind conict
deaths as in SSA, where the regional percentage of conict
deaths was much greater than that of terrorism-related
casualties. Because insurgencies often involve some terrorist
tactics, the regional shares for casualties can be more closely
tied for conicts and terrorism.
Figure 5 provides an annual perspective of conict deaths
since the start of 1989, based on UCDP data (Sundberg and
Melander, 2013). The overall impression is that these conict
deaths were mainly tied to SSA, SAS, and MENA throughout
the sample period, with some link to ECA until around 2000.
In Figure 5, the two highest SSA peaks were driven by the
Ethiopia-Eritrea conicts. The persistent deaths in MENA
since 2014 were because of IS and civil wars in Yemen, Syr-
ia, and Libya. Insurgencies in Sri Lanka and Afghanistan
were behind recurrent peaks in SAS.
Our last conict perspective involves the individual MENA
countries, 13 of which surpassed the conict threshold of 25
battle-related deaths once or more times during 19892018
based on UCDP/PRIO (International Peace Research Institute,
Oslo) Armed Conict Dataset Version 19.1 (Gleditsch et al.,
2002; Pettersson et al., 2019). For the two sample periods,
Figure 6 indicates the number of conicts of a MENA coun-
try. Countries with no conicts are not listed. A country may
experience multiple conicts having the requisite deaths but
with different antagonists in a given calendar year, for
example, Israel battled both Fatah and Hezbollah in 1990
resulting in two conicts that year. Throughout 19892018,
the MENA countries most plagued by conicts were Israel,
Iraq, Algeria, Iran, Syria, and Egypt. For the most conict-
prone countries during the pre-2002 period, the conict
threshold was met or surpassed the following times: 20,
Israel; 14, Iraq; 11, Algeria; 9, Iran; 6, Egypt; and 5, Djibouti.
During the post-2002 period, the conict threshold of con-
ict-ridden MENA countries was met or surpassed as follows:
Figure 4. Distribution of deaths by conicts for two time periods.
Global Policy (2020) ©2020 The Authors. Global Policy published by Durham University and John Wiley & Sons Ltd.
Middle East and North Africa 7
18, Algeria; 16, Syria; 15, Israel; 15, Iraq; 12, Iran; 11, Yemen;
8, Libya; and 7, Egypt. MENA is a region whose tranquility
was disrupted in eight countries by civil wars, insurgencies,
or terrorist campaigns since 1989. In addition, conicts
shifted recently to embroil Yemen, Syria, and Libya in civil
wars. The Gulf Wars and the Arab Spring have fueled contin-
ued regional instability and conicts, as unintended conse-
Theoretical and empirical considerations in the
study of global and MENA terrorism
Theoretical perspective on the drivers of terrorism
Through a series of panel regressions for 19702018 and
a recent subperiod, we seek to identify the main
determinants of domestic and transnational terrorist inci-
dents for global and MENA samples. In so doing, our
dependent variable is either the number of domestic or
transnational terrorist attacks for the relevant sample and
time periods. There is a rich literature on the determi-
nants of terrorism (e.g., Gaibulloev and Sandler, 2019;
Gassebner and Luechinger, 2011; Krieger and Meierrieks,
2011). We draw from this literature to assemble relevant
independent variables, commencing with the global analy-
sis before considering MENA.
A countrys population (POP) provides more potential tar-
gets for terrorist attacks, offers a greater terrorist recruitment
pool, and allows for more cover. As such, population is
anticipated to exert a positive inuence on either domestic
or transnational terrorist incidents (e.g., Campos and Gasseb-
ner, 2013; Piazza, 2006). As is customary, a log
Figure 5. The number of deaths by conicts, 19892018.
1989 1993 1997 2001 2005 2009 2013 2017
Figure 6. The number of conicts for two time periods in MENA.
0 5 10 15 20 25 30 35 40
1989-2001 2002-2018
©2020 The Authors. Global Policy published by Durham University and John Wiley & Sons Ltd. Global Policy (2020)
Wukki Kim and Todd Sandler
transformation is applied to this variable. The log of gross
domestic product per capita (GDP/POP) may have a positive
or negative inuence on the two forms of terrorism. During
the reign of the leftist terrorist up through the early 1990s,
a positive inuence is predicted because these leftists
resided in afuent countries (Abadie, 2006; Piazza, 2006). If,
instead, poverty drives terrorism, then a negative relation-
ship between GDP/POP and alternative forms of terrorism is
expected. However, the literature usually nds a positive
empirical relationship, which is against the conventional
poverty argument (Enders et al., 2016; Gassebner and Lue-
chinger, 2011). Democracy is viewed as exerting opposing
effects on terrorism. This regime type can facilitate terrorism
through strategic considerations including freedoms of
movement, association, and privacy. Constraints on the
countrys executive branch (e.g., preservation of due pro-
cess) can further foster terrorism. Also, press freedoms can
augment terrorism in democracies as terrorists get much-
needed visibility (Eubank and Weinberg, 1994; Li, 2005). By
contrast, democracy-induced political participation can limit
grievances and, thus, terrorism (Eyerman, 1998). By abiding
by its mandate to protect lives, democracies can also limit
terrorism through swift counterterrorism measures. Autocra-
cies can also take draconian actions against terrorists to
curb terrorism. Gaibulloev et al. (2017) argue that these
opposing inuences mean that highly democratic and
highly autocratic states have relatively low levels of terror-
ism compared to in-between anocratic regimes that have
greater difculty in either protecting lives or assuaging grie-
vances. Anocracies often characterize autocracies that are
transitioning to democracy (i.e., some MENA regimes after
the Arab Spring) and do not yet have sufciently function-
ing bureaucratic institutions. Gaibulloev et al. (2017) predict
an inverted U-shaped relationship between regime type and
terrorism so that anocracies are most plagued by terrorism
(also see Lai, 2007). This relationship can be tested with a
quadratic measure of regime type based on a regime index
and its squared value.
At the global level, six regions (EAP, ECA, LAC, NA, SAS,
and SSA) are anticipated to experience less terrorism than
MENA during 19702018 as hinted by earlier-presented Fig-
ures 1 and 2. The number of civil conicts should fuel ter-
rorist attacks with a larger positive effect on domestic,
relative to transnational, terrorism. As a deterrent, a coun-
trys share of GDP devoted to defense is expected to have a
negative inuence on both kinds of terrorism. If, by contrast,
greater defense spending augments grievances with heavy-
handed actions, the defense share may be a positive driver
of terrorism. Finally, the post-9/11 era should reduce
transnational terrorism as borders are tightened; but this era
may increase domestic terrorism, especially in hotspots like
MENA, as terrorist attacks are not exported.
There are a few necessary adjustments when ascertaining
the determinants of MENA terrorism. Given the signicant
reduction in sample size, we must reduce the number of
independent variables to conserve degrees of freedom.
Hence, we replace our two measures of democracy with a
single measure, Democracy, that distinguishes between
democratic and non-democratic states. In so doing, we do
not include a regime index and its squared value (described
in the following section) to test for a nonlinear relationship.
Also, we include a dummy for the post-Arab Spring era of
20102018, which we expect increases MENA terrorism. For
the MENA estimates, there is obviously no need for regional
dummies. In a robustness test, we include the number of
religious fundamentalist terrorist groups in the MENA
In our global sample for 19702018, there are 164 coun-
tries that are regionally distributed as follows: 19, EAP; 48,
ECA; 23, LAC; 2, NA; 7, SAS; 45, SSA, and 20, MENA. GDP/
POP,POP, and Defense shares of GDP are drawn from the
World Bank (2019), along with regional classications.
Regime or Polity scores vary from 10 (highly autocratic)
to 10 (highly democratic) (Marshall et al., 2018). Polity val-
ues below 5 denote autocracies, values 6 and above indi-
cate democracies, and values from 5 to 5 correspond to
The total number of terrorist incidents comes
from START (2019) for which we apply the methods of
Enders et al. (2011) to distinguish between domestic and
transnational attacks. The number of civil conicts is drawn
from UCDP/PRIO Armed Conict Dataset (Gleditsch et al.,
2002; Pettersson et al., 2019). For robustness runs, the
number of religious terrorist groups in MENA countries is
taken from Hou et al. (2020). A larger number of such
groups is anticipated to fuel terrorism, especially at the
domestic level. Additionally, we use a measure of eco-
nomic freedom, which averages ve indicators that account
for government size, legal system/property rights, sound
money, trade freedom, and regulation (The Fraser Institute,
2019). This economic freedom measure is reported from
1970 until 2000 at ve-year intervals; thereafter, it is
reported annually.
Table 2 indicates our independent variables, their sources,
mean values, standard deviation (SD), and minimum (Min) and
maximum (Max) values over the panel. By taking the anti-loga-
rithm of population, we derive a mean value of 34.05 million
persons for an average country-year. The Min is 0.054 million
persons for the Seychelles in 1970 and the Max is 1.392 billion
persons for China in 2018. The converted mean for GDP/POP is
US$11,107 averaged over all countries and years, with a Min of
US$161.7 for Myanmar in 1973 and a Max of US$116,232.8 in
UAE in 1980. The mean Defense shares of GDP is 2.79 per cent
for the entire sample period, while it is 3.65 per cent, 2.96 per
cent, and 1.99 per cent for 19701989, 19902001, and 2002
2018, respectively, indicating a decline since the end of the
Cold War. The average Polity score is 1.41, indicative of an anoc-
racy. For the Democracy dummy, 44 per cent of our sample has
aPolity score 6 or higher on average. The regional mean indi-
cates each regions average share of the sample e.g., EAP
accounts for 12 per cent of sample countries during 1970
2018. For sample countries, the average annual number of
transnational and domestic terrorist attacks is 1.8 and 13.33,
respectively. Germany and the United Kingdom experienced
Global Policy (2020) ©2020 The Authors. Global Policy published by Durham University and John Wiley & Sons Ltd.
Middle East and North Africa 9
the largest number of transnational attacks of 135 in 1995 and
2013, respectively, while Iraq experienced the greatest number
of domestic attacks of 3,098 in 2014. The average number of
civil conicts per country-year was 0.25 with a Max of 7 corre-
sponding to India in 1996, 1998, 1999, 2000, and 2003. More-
over, the average annual number of religious terrorist groups is
0.45 with Pakistan having the most 27 in 2015. Finally, the
mean economic freedom index is 6.08 out of 10 with Singapore
displaying the largest index of 8.82 during 19951999.
Empirical methodology
Given that our dependent variable is the count of domestic
or transnational terrorist incidents, we must use a regression
whose distribution is geared to count data. Options include
Poisson or negative binomial (NB) regressions. If there is
overdispersion such that the observed variance exceeds the
sample mean, then the NB regression is appropriate
(Cameron and Trivedi, 2013). When we later test for the
underlying distribution in terms of Ln (alpha), overdispersion
is conrmed for all our regressions at the .01 level of signi-
cance. Thus, we display NB estimates throughout the text
and online appendix. Moreover, we report venue-country-
clustered robust standard errors.
In Table 3, the NB regressions identify potential determinants
for domestic and transnational terrorist incidents for 1970
2018 and 20022018. We begin with the results for domestic
attacks during the entire sample period. Population has the
anticipated positive inuence, where a 1 per cent increase in a
sample countrys population raises the countrys expected
number of domestic incidents by 0.73 per cent after
transforming the coefcient appropriately.
For the entire per-
iod, a 1 per cent increase in GDP per capita augments expected
domestic terrorist incidents by 0.20 per cent, which suggests
that poverty is not a driver of such attacks. The signicant posi-
tive coefcient on Polity and the signicant negative coef-
cient on Polity squared are consistent with an inverted U-
shaped relationship for which most domestic terrorism occurs
in anocracies.
Next, we consider the inuence of the six regio-
nal dummies relative to MENA. In the case of EAP, domestic ter-
rorism is 82.5 per cent smaller than that in MENA.
terrorism is also signicantly smaller in ECA, NA, and SSA by
54.5 per cent, 84.5 per cent, and 64.7 per cent, respectively,
compared to MENA. The signicant positive coefcient on the
Number of civil conicts means that domestic terrorism
increases by about 322 per cent for every additional internal
conict. This represents the expected strong relationship
between such conicts and domestic terrorism, consistent with
our overview of such conicts in Figures 46. The Defense
shares of GDP has a positive inuence where a 1 per cent
increase in these shares raises domestic terrorism by 13.2 per
cent, so that defense is not having a deterrent effect. Finally,
the post-9/11 era does not experience a signicant fall in
domestic terrorism.
Next, we consider domestic terrorism estimates for 2002
2018. Domestic terrorism is still positively correlated with
Population. However, income per capita and regime type are
no longer signicant determinants of domestic attacks in
this much-abbreviated period. Five of the six regions the
exception being SAS experience signicantly less domestic
terrorism than MENA, consistent with Figures 1 and 2. The
Number of civil conicts greatly raises domestic terrorism for
which an additional conict increases this terrorism by 432.8
per cent. Finally, Defense shares of GDP only marginally
increases domestic terrorism.
Table 2. Summary statistics
Count Mean SD Min Max Source
Ln (Population) 7957 15.86 1.65 10.89 21.06 World Bank (2019)
Ln (GDP per capita) 6794 8.25 1.55 5.09 11.66 World Bank (2019)
Defense shares of GDP 6161 2.79 3.33 0 117.35 World Bank (2019)
Polity score 6995 1.41 7.29 10 10 Marshall et al. (2018)
Democracy 6995 0.44 0.50 0 1 Marshall et al. (2018)
East Asia and Pacic 8036 0.12 0.32 0 1 World Bank (2019)
Europe and Central Asia 8036 0.29 0.46 0 1 World Bank (2019)
Latin America and Caribbean 8036 0.14 0.35 0 1 World Bank (2019)
Middle East and North Africa 8036 0.12 0.33 0 1 World Bank (2019)
North America 8036 0.01 0.11 0 1 World Bank (2019)
South Asia 8036 0.04 0.20 0 1 World Bank (2019)
Sub-Saharan Africa 8036 0.27 0.44 0 1 World Bank (2019)
Num of transnational incidents 8036 1.80 7.01 0 135 GTD (2019) and Enders et al. (2011)
Num of domestic incidents 8036 13.33 87.94 0 3098 GTD (2019)
Num of civil conicts 8036 0.25 0.67 0 7 UCDP/PRIO Armed Conict Dataset version.19.1
Num of religious groups 8036 0.45 1.67 0 27 Hou et al. (2020)
Economic freedom index 5188 6.08 1.90 1.84 8.82 The Fraser Institute (2019)
©2020 The Authors. Global Policy published by Durham University and John Wiley & Sons Ltd. Global Policy (2020)
Wukki Kim and Todd Sandler
We now consider the drivers of transnational terrorist
incidents for 19702018. Generally, the estimates are like
those for domestic terrorism for this period. Population,
GDP per capita,Number of civil conicts, and Defense shares
of GDP are positive inuences on transnational attacks.
Also, most transnational terrorism is associated with anoc-
racies. EAP, NA, and SSA display signicantly fewer transna-
tional terrorist incidents relative to MENA. The Number of
civil conicts is a smaller positive determinant of transna-
tional terrorism attacks one additional internal conict
raises terrorism by 149.2 per cent which is less than half
the inuence on domestic terrorism during 19702018.
Some consequences of these conicts arise owing to cross-
border concerns and international intervention. Another
notable difference is that transnational terrorism greatly
falls during the post-9/11 era, where the 0.905 coefcient
is consistent with a 59.6 per cent drop, which likely stems
from more secure borders.
For 20022018, the determinants of transnational terror-
ism are quite like those for the entire period, except where
noted. Now ve of the six regions display less transnational
terrorism relative to MENA, thereby highlighting the key
post-9/11 role that MENA plays as a driver of transnational
terrorism. The Number of civil conicts displays a much
diminished, but still important, inuence on transnational,
relative to domestic, terrorism.
In Table 4, overdispersion is again detected so that our
four sets of estimates for the MENA sample apply the NB
estimator. As explained earlier, there are fewer independent
variables for the MENA runs. For 19702018, domestic ter-
rorism rises with increases in Population, democratic
regimes, the Number of civil conicts, and the post-9/11 era.
MENAs democracies have almost four times as much
domestic terrorism as their non-democratic counterparts.
Since 2002, Iraq, Lebanon, and Tunisia moved from anocracy
to democracy, while Egypt, Libya, and Morocco moved from
Table 3. Determinants of domestic or transnational terrorist incidents in global sample, based on two different time periods
Domestic incidents Transnational incidents
19702018 20022018 19702018 20022018
Coefcients Coefcients Coefcients Coefcients
Ln (Population) 0.729*** 0.717*** 0.677*** 0.793***
(0.084) (0.090) (0.079) (0.153)
Ln (GDP per capita) 0.257** 0.198 0.193* 0.233
(0.101) (0.130) (0.108) (0.176)
Polity score 7.042*** 3.289 6.924*** 7.874***
(2.101) (3.200) (1.345) (2.544)
Polity score squared 5.810*** 2.086 5.610*** 5.799***
(1.850) (2.743) (1.258) (2.177)
East Asia and Pacic1.743*** 2.052*** 1.578*** 2.222***
(0.505) (0.681) (0.344) (0.449)
Europe and Central Asia 0.788* 1.785*** 0.015 0.801**
(0.467) (0.502) (0.371) (0.403)
Latin America and Caribbean 0.327 1.935*** 0.314 2.068***
(0.467) (0.507) (0.335) (0.444)
North America 1.864*** 2.875*** 1.398*** 4.114***
(0.524) (0.632) (0.370) (1.082)
South Asia 0.282 0.819 0.545 0.020
(0.511) (0.564) (0.477) (0.622)
Sub-Saharan Africa 1.042*** 1.213*** 0.732*** 0.589**
(0.370) (0.423) (0.276) (0.239)
Number of civil conicts 1.440*** 1.673*** 0.913*** 0.791***
(0.254) (0.314) (0.225) (0.252)
Defense shares of GDP 0.124** 0.150* 0.078** 0.193***
(0.057) (0.081) (0.037) (0.058)
Time (20022018) 0.100 0.905***
(0.206) (0.224)
Constant 14.142*** 12.704*** 13.912*** 17.451***
(1.678) (2.043) (1.812) (3.758)
Ln (alpha) 1.641*** 1.576*** 1.279*** 1.463***
(0.073) (0.094) (0.087) (0.188)
N 5,613 2,364 5,613 2,364
Notes:: Cluster-robust standard errors based on country are in the parentheses.
*p<0.1,;**p<0.05, and; ***p<0.01.
Global Policy (2020) ©2020 The Authors. Global Policy published by Durham University and John Wiley & Sons Ltd.
Middle East and North Africa 11
autocracy to anocracy. Six of 20 MENA countries became
more democratic. Algeria, Djibouti, Jordan, and Yemen
stayed anocracies. Israel maintained its democracy, while the
other MENA countries are autocracies. Transitioning regimes
are particularly prone to terrorism. Based on the empirical
ndings in Table 4, the policy of encouraging democracies
in this region was clearly unwise in regard to domestic ter-
rorism. Internal wars have a large effect on domestic terror-
ism one additional war increases the number of domestic
terrorist incidents by almost nine-fold. The post-9/11 era
sustains 2.48 times as many domestic attacks as the pre-9/
11 period. The Defense shares of GDP has the anticipated
deterrent effect.
For 20022018, the drivers of domestic terrorism in MENA
are analogous to those for the entire period, except for
defense spending and the Arab Spring. The inuence of
Democracy on domestic terrorism rises sharply from 396.8 per
cent to 736.5 per cent relative to non-democratic regimes
when the entire period is compared to 20022018. The Arab
Spring era experiences a huge increase in domestic terrorism
of 2,358.2 per cent. The negative inuence of democracy on
domestic terrorism is due to some edgling democracies that
do not possess strong bureaucracies to support governance
or to address grievances (Gaibulloev et al., 2017).
For transnational terrorism, there are a few noteworthy
differences compared to corresponding MENA estimates for
domestic terrorism. Internal wars drive transnational terror-
ism but at a much lower rate than domestic terrorism for
corresponding periods. Population is only marginally
signicant in inducing more transnational attacks during
19702018. Democracy still encourages transnational terror-
ism but at a reduced rate of 130.9 per cent (at the 0.10
level) and 288.1 per cent during 19702018 and 20022018,
respectively. Defense spending positively affects transna-
tional terrorism during 20022018. The effect of the Arab
Spring is a much-reduced rate of 138.9 per cent for transna-
tional terrorism, compared to domestic terrorism.
In the online Sppendix, we provide 16 sets of robustness
runs. In Table A1, we display robustness runs for domestic
or transnational terrorism for the global sample, where we
include two new variables Economic freedom index and
Number of religious groups. By promoting trade, property
rights, and deregulation, economic freedom is expected to
promote prosperity and reduce grievances, thereby curtail-
ing both forms of terrorism (Basuchoudhary and Shughart,
2010). In contrast, the prevalence of religious fundamentalist
terrorist groups is anticipated to increase both types of ter-
rorism. Because panel data on religious terrorist groups are
only available until 2016, we must truncate the estimates by
two years. Qualitatively, the coefcient estimates in Table A1
are amazingly consistent with those of analogous variables
in Table 3. For the global sample, only the Number of reli-
gious groups displays the anticipated positive inuence on
transnational terrorism for 19702016. Economic freedom
does not show any signicant impact.
Table 4. Determinants of domestic or transnational terrorist incidents in MENA, based on two different time periods
Domestic incidents Transnational incidents
19702018 20022018 19702018 20022018
Coefcients Coefcients Coefcients Coefcients
Ln (Population) 0.511*** 0.493** 0.299* 0.291
(0.165) (0.245) (0.178) (0.204)
Ln (GDP per capita) 0.324 0.268 0.023 0.074
(0.394) (0.440) (0.245) (0.299)
Democracy 1.603*** 2.124*** 0.837* 1.356***
(0.460) (0.402) (0.474) (0.481)
Number of civil conicts 2.278*** 2.715*** 0.942** 1.186***
(0.599) (0.539) (0.390) (0.398)
Defense shares of GDP 0.105** 0.018 0.019 0.153**
(0.053) (0.085) (0.049) (0.068)
Time (20022018) 1.247** 0.366
(0.511) (0.490)
Time (20102018) 3.202*** 0.871**
(0.464) (0.353)
Constant 10.237** 11.658* 4.959 5.723
(5.041) (7.050) (4.444) (5.490)
Ln (alpha) 1.828*** 1.315*** 1.325*** 1.029***
(0.245) (0.211) (0.269) (0.268)
N 594 254 594 254
Notes:: Cluster-robust standard errors based on country are in the parentheses.
*p<0.1,;**p<0.05, and; ***p<0.01.
©2020 The Authors. Global Policy published by Durham University and John Wiley & Sons Ltd. Global Policy (2020)
Wukki Kim and Todd Sandler
Table A2 augments our four MENA runs to include the
economic freedom and religious groups variables. Key inde-
pendent variables Democracy,Number of civil conicts, and
the Arab Spring era provide similar ndings to those in
Table 4. A noteworthy difference is that internal wars have a
smaller, but still signicant, positive impact on the two
forms of terrorism for the sample periods. In addition, the
Arab Spring era is only a signicant inducer of domestic ter-
rorism in MENA. Economic freedom now has the predicted
negative and signicant effect on domestic terrorism. More-
over, the prevalence of religious terrorist groups increases
domestic and transnational terrorism in MENA for both sam-
ple periods, which is notably different than for the global
For a global sample, another set of robustness checks,
reported in online Table A3, investigates the determinants
of domestic and transnational terrorist incidents for two ear-
lier periods: 19701989 and 19902001. During the earlier
intervals, both types of terrorism still indicate a signicant
positive response to population and internal conicts. Once
again, an inverted U-shaped response characterizes regime
type with anocracies attracting the most terrorism. Given
the prevalence of terrorism in LAC and ECA during 1970
1989 (see Figures 1 and 2), the larger response in these two
regions relative to MENA is to be expected. For the brief
19902001 period, population and internal conicts con-
tinue to be positive drivers of both forms of terrorism. The
regime type is not a signicant inuence probably due to
the periods brevity. Domestic terrorism in EAP and NA fall
relative to that in MENA. For transnational terrorism, NA
attacks decline greatly compared to MENA, while LAC
attacks rise relative to those in MENA.
Table A4 complements Table 4 as we examine two earlier
hotspots for terrorism LAC and ECA for the 19702001
period. By necessity, we drop the post-2001 time dummies
in Table 4 for this new pre-2002 exercise. We again nd
Population, the Number of civil conicts, and Defense share of
GDP to be robust positive determinants of domestic and
transnational terrorism in these two earlier regional hot-
spots. The noteworthy distinction is that, unlike MENA,
Democracy is not a determinant of terrorism in LAC and ECA
during 19702001.
Concluding remarks and policy implications
Given its strategic location and oil wealth, MENA plays a piv-
otal role in world stability and economic prosperity. The spil-
lover of terrorism to Europe and North America, coupled with
the recent refugee exigency in Europe stemming from three
civil wars in MENA, aptly illustrate the regions impact on
political stability and violence. The papers comparative study
of terrorism and conicts in MENA and the other six regions
point to the following signicant ndings. The regional distri-
bution of terrorism is not static but affected by major events
(e.g., the rise of religious fundamentalist terrorists after 1990)
and policy reactions to these events (e.g., enhanced border
security following 9/11). Since 1989, there have been large-
scale geographical displacements of terrorism to MENA, SAS,
and SSA away from LAC and ECA. For policy purposes, these
displacements require repositioning of counterterrorism
assets (e.g., intelligence ofcers, conict assistance, and com-
mando squads). The analysis here also underscores the rising
importance of domestic terrorism over transnational terrorism
for all seven regions, especially MENA. With this shifting
importance, developed-country donors are less inclined to
supply counterterrorism assistance because their assets are
not so threatened at home or abroad. This response is proba-
bly short-sighted because domestic terrorism is shown else-
where to induce more transnational terrorism over time as
terrorists try to gain more attention after honing their craft
(Enders et al., 2011). For 20022012, MENA now accounts for
almost half of all terrorism casualties, making this region the
epicenter of global terrorism. This is an alarming statistic for
world security against terrorism. From our country-by-country
breakdown, terrorism is mostly prevalent since 2002 in six
MENA countries involved in civil wars or confronting insur-
gencies, thereby pinpointing where any aid against terrorism
should ow.
From a conict viewpoint, SAS, SSA, and MENA now
account for the lions share of conict deaths worldwide.
The correlation between terrorist violence and conicts
strongly suggests that efforts to curb conicts through
reduced arms sales and peacekeeping missions may pay a
double dividendby also ameliorating terrorism. The geo-
graphical distribution of conicts morphed over time in the
same direction as terrorism. Given conicts in SAS, SSA, and
MENA, NATO must contemplate a greater peacekeeping
presence in these regions; however, NATOs decreasing
coherence under the Trump administration sties such a
change. Currently, the United Nations deployed peacekeep-
ers to Africa and Asia, while NATO dispatched peacekeepers
specically to Iraq and Afghanistan. Even the latter is being
drawn down under the current US administration. The three
civil wars in Libya, Syria, and Yemen foment regional insta-
bility with negative spillovers to Europe. For Syria and
Yemen, arms sales by major arms suppliers, including the
United States and Russia, foster the carnage. Without greater
restraint by these suppliers, total military spending in MENA
is poised to rise sharply and may spark future conicts in
this volatile region, fueled by religious hatreds. Actions to
assist countries in MENA to develop their own arms indus-
tries, which Saudi Arabia and others seek to do, would elimi-
nate current controls by arms suppliers to constrain client
nationshostile actions. This development should be resisted
to bolster world peace.
Anal policy conclusion is that pushing democracy in
MENA is not a useful counterterrorism policy unless there
are enhanced efforts to provide supporting infrastructure
and institutions, coupled with the elimination of civil con-
icts. The introduction of democracy needs to be more
gradual and thoughtful.
Data availability statement
Replication les at:
Global Policy (2020) ©2020 The Authors. Global Policy published by Durham University and John Wiley & Sons Ltd.
Middle East and North Africa 13
1. These terrorist attack percentages are derived from START (2019)
Global Terrorism Database (GTD).
2. On the resource curse, consult Ali & Abdellatif (2015) and Ross
3. Every terrorist event data set suffers from some coding inconsis-
tency and GTD is no exception (Enders et al., 2011). GTD is more
consistently coded from 1990 after being taken over by START.
Given the rather stark changes depicted in Figures 13 for the three
time periods, we do not envision inconsistent coding to change our
main takeaways.
4. The share of domestic terrorist attacks equals domestic attacks as a
proportion of domestic plus transnational attacks. Hence, the share
of transnational terrorist attacks equals one minus the share of
domestic terrorist attacks.
5. The relatively high share of transnational terrorism in ECA was due,
in part, to attacks by IS, al-Qaida, Kurdistan WorkersParty (PKK),
and Chechen rebels during the post-9/11 period.
6. Throughout this section, the attributed conicts along with death
tolls are drawn from Sundberg & Melander (2013).
7. We note that Drakos & Gofas (2006) caution about the possible
underreporting of terrorism in autocracies. This is less apt to be a
problem for MENA where there is a large foreign press presence.
8. Given our NB estimation, this transformation equals: exp[0.729 9ln
(1.01)] 1=1.00728 1=0.00728 Other ln coefcients are trans-
lated in an analogous fashion.
9. During 20022018, the insignicance for the inverted U-shaped rela-
tionship for regime type for domestic incidents is likely due to the
relatively short interval of time.
10. This follows from exp(1.743) 1=0.825 =82.5%.
11. This follows from exp(1.603)1=3.9679 =396.79%.
Abadie, A. (2006) Poverty, Political Freedom and the Roots of
Terrorism,American Economic Review, 96 (2), pp. 5056.
Ali, H. E. and Abdellatif, O. A. (2015) Military Expenditures and Natural
Resources: Evidence from Rentier States in the Middle East and
North Africa,Defence and Peace Economics, 26 (1), pp. 513.
Bandyopadhyay, S., Sandler, T. and Younas, J. (2011) Foreign Aid as
Counterterrorism Policy,Oxford Economic Papers, 63 (3), pp. 423
Basuchoudhary, A. and Shughart, W. F. (2010) On Ethnic Conict and
the Origins of Transnational Terrorism,Defence and Peace Economics,
21 (1), pp. 6587.
Cameron, A. C. and Trivedi, P. K. (2013) Regression Analysis of Count
Data. New York: Cambridge University Press.
Campos, N. F. and Gassebner, M. (2013) International Terrorism,
Domestic Political Instability and the Escalation Effect,Economics &
Politics, 25 (1), pp. 2747.
Drakos, K. and Gofas, A. (2006) The Devil You Know but Are Afraid to
Face: Underreporting Bias and Its Distorting Effects on the Study of
Terrorism,Journal of Conict Resolution, 50 (5), pp. 714735.
Enders, W. and Sandler, T. (1993) The Effectiveness of Antiterrorism
Policies: A Vector-Autoregression-Intervention Analysis,American
Political Science Review, 87 (4), pp. 829844.
Enders, W. and Sandler, T. (1999) Transnational Terrorism in the Post-
Cold War Era,International Studies Quarterly, 43 (2), pp. 145167.
Enders, W. and Sandler, T. (2000) Is Transnational Terrorism Becoming
More Threatening? A Time-series Investigation,Journal of Conict
Resolution, 44 (3), pp. 307332.
Enders, W. and Sandler, T. (2006) Distribution of Transnational Terrorism
among Countries by Income Classes and Geography after 9/11,
International Studies Quarterly, 50 (2), pp. 367393.
Enders, W. and Sandler, T. (2012) The Political Economy of Terrorism, 2nd
edn. Cambridge: Cambridge University Press.
Enders, W., Hoover, G. A. and Sandler, T. (2016) The Changing
Nonlinear Relationship between Income and Terrorism,Journal of
Conict Resolution, 60 (2), pp. 195225.
Enders, W., Sandler, T. and Gaibulloev, K. (2011) Domestic versus
Transnational Terrorism: Data, Decomposition, and Dynamics,
Journal of Peace Research, 48 (3), pp. 319337.
Eubank, W. L. and Weinberg, L. B. (1994) Does Democracy Encourage
Terrorism?,Terrorism and Political Violence, 6 (4), pp. 417435.
Eyerman, J. (1998) Terrorism and Democratic States: Soft Targets or
Accessible Systems?,International Interactions, 24 (2), pp. 151170.
Findley, M. E. and Young, J. K. (2012) Terrorism and Civil War: A Spatial
and Temporal Approach to a Conceptual Problem,Perspectives on
Politics, 10 (2), pp. 285305.
Gaibulloev, K. and Sandler, T. (2019) What We Have Learned about
Terrorism since 9/11,Journal of Economic Literature, 57 (2), pp. 275
Gaibulloev, K., Sandler, T. and Santifort, C. (2012) Assessing the Evolving
Threat of Terrorism,Global Policy, 3 (2), pp. 135144.
Gaibulloev, K., Piazza, J. A. and Sandler, T. (2017) Regime Types and
Terrorism,International Organization, 71 (3), pp. 491522.
Gassebner, M. and Luechinger, S. (2011) Lock, Stock, and Barrel: A
Comprehensive Assessment of the Determinants of Terror,Public
Choice, 149 (34), pp. 235261.
Gleditsch, N. P., Wallensteen, P., Eriksson, M., Sollenberg, M. and Strand,
H. (2002) Armed Conict 19462001: A New Dataset,Journal of
Peace Research, 39 (5), pp. 615637.
Hoffman, B. (2006) Inside Terrorism: Revised and Expanded Edition. New
York: Columbia University Press.
Hou, D., Gaibulloev, K. and Sandler, T. (2020) Introducing Extended Data
on Terrorist Groups (EDTG), 1970 to 2016,Journal of Conict
Resolution, 64 (1), pp. 199225.
Krieger, T. and Meierrieks, D. (2011) What Causes Terrorism?,Public
Choice, 147 (12), pp. 327.
Lai, B. (2007) Draining the Swamp: An Empirical Examination of the
Production of International Terrorism,Conict Management and
Peace Science, 24 (4), pp. 297310.
Li, Q. (2005) Does Democracy Promote Transnational Terrorist
Incidents?,Journal of Conict Resolution, 49 (2), pp. 278297.
Marshall, M. G., Jaggers, K. and Gurr, T. R. (2018) Polity IV Project:
Dataset and UsersManual. Vienna, VA: Center for Systemic Peace:
Polity IV Project [online]. Available from: http://www.systemicpeace.
org/inscrdata.html [Accessed 23 September 2019].
Mickolus, E. F., Sandler, T., Murdock, J. M. and Flemming, P. A. (2019)
International Terrorism: Attributes of Terrorist Events, 19682018
(ITERATE). Ponte Vedra: Vinyard Software.
National Consortium for the Study of Terrorism and Responses to Terrorism
(START) (2019) Global Terrorism Database(GTD) [online]. Available
from: [Accessed 7 October 2019].
Pettersson, T., H
ogbladh, S. and
Oberg, M. (2019) Organized Violence,
19892018 and Peace Agreements,Journal of Peace Research, 56 (4),
pp. 589603.
Piazza, J. A. (2006) Rooted in Poverty? Terrorism, Poor Economic
Development, and Social Cleavages,Terrorism and Political Violence,
18 (1), pp. 159177.
Piazza, J. A. (2007) Draining the Swamp: Democracy Promotion, State
Failure, and Terrorism in 19 Middle Eastern Countries,Studies in
Conict & Terrorism, 30 (6), pp. 521539.
Rapoport, D. C. (2004) The Four Waves of Modern Terrorism,inA.K.
Cronin and J. Ludes (eds.), Attacking Terrorism: Elements of a Grand
Strategy. Washington: Georgetown University Press, pp. 4673.
Ross, M. L. (2004) What Do We Know about Natural Resources and Civil
War?,Journal of Peace Research, 41 (3), pp. 337356.
Sandler, T. and George, J. (2016) Military Expenditure Trends for 1960
2014 and What They Reveal,Global Policy, 7 (2), pp. 174184.
©2020 The Authors. Global Policy published by Durham University and John Wiley & Sons Ltd. Global Policy (2020)
Wukki Kim and Todd Sandler
SIPRI (2019) SIPRI Yearbook 2019: Armaments, Disarmaments and
International Security. Oxford: Oxford University Press.
Stockholm International Peace Research Institute (SIPRI) (2017) SIPRI
Yearbook 2017: Armaments, Disarmaments and International Security.
Oxford: Oxford University Press.
Sundberg, R. and Melander, E. (2013) Introducing the UCDP
Georeferenced Event Dataset,Journal of Peace Research, 50 (4), pp.
The Fraser Institute (2019) Data from Economic Freedom of the World.
Available from:
freedom [Accessed 15 February 2020].
World Bank (2019) World Development Indicators[online]. Available
from: [Accessed 4
November 2019].
Supporting Information
Additional supporting information may be found online in the Supporting
Information section at the end of the article.
Table A1. Determinants of domestic or transnational terrorist incidents in
global sample (robustness).
Table A2. Determinants of domestic or transnational terrorist incidents in
MENA (robustness).
Table A3. Determinants of terrorist incidents in global sample, based on
two different early time periods.
Table A4. Determinants of domestic or transnational terrorist incidents in
LAC and ECA during 19702001
Author Information
Wukki Kim is an assistant professor in the Department of Economics
and Law at the Korea Military Academy, Seoul, Republic of Korea. He
holds the rank of Major in the Korean armed forces and earned his PhD
at the University of Texas at Dallas in April 2020. His research interests
are in defense economics, peacekeeping, and terrorism.
Todd Sandler is the Vibhooti Shukla Professor of Economics and Politi-
cal Economy at the University of Texas at Dallas. His research interests
include collective action, environmental economics, terrorism, public
economics, and defense economics. His most recent book is Terrorism:
What Everyone Needs to Know (Oxford University Press, 2019).
Global Policy (2020) ©2020 The Authors. Global Policy published by Durham University and John Wiley & Sons Ltd.
Middle East and North Africa 15
... We utilise unique data for banks located within the Middle East and North Africa (MENA) region. Because of excessive and congested regional conflicts, political violence and vast oil wealth, MENA plays a key role in the prospects for global terrorism, world tranquillity and economic prosperity (Kim and Sandler 2020). Therefore, the choice of MENA countries in our study is justified by the large number of terrorist attacks over the last two decades (Bardwell and Iqbal 2021). ...
... According to the Global Terrorism Index (2019), between 2002 and 2018, the MENA region recorded the highest number of fatalities due to terrorism (93,700, representing 42% of the global total). In fact, the MENA region is constantly suffering from recurring terrorist attacks in addition to political unrest, which reflects the importance of understanding the impact of terrorism on bank stability in this region (Kim and Sandler 2020). 2 We employ different risk indicators (i.e., insolvency, credit, liquidity, asset and operational risks) and performance measures (i.e., profitability ratio and cost to income ratio). ...
... Hence, our work contributes to the growing banking studies which concentrate on measuring economic stability and risk mitigation, but which have not captured the individual and joint effects of terrorist attacks on financial performance and bank risk (e.g., Gasbarro et al. 2002;Imerman 2020;Trinh et al. 2020b;Elnahass et al. 2021). In fact, our results (e.g., highlighting the damaging effect of terrorism on bank risk) extend earlier work on the economic impacts of terrorism (Estrada et al. 2015;Asongu and Nwachukwu 2017a, b;Kim and Sandler 2020). This study is also of relevance to investment choices in the emerging banking industry in general (e.g., Rahman and Hassan 2013;Fosu et al. 2020), and studies examining financial determinants of bank stability and corporate policies in monarchies in particular (Choi 2010). ...
Full-text available
This study examines the impact of terrorism on bank stability, represented by bank risk and financial performance. We consider banks from 14 countries located in the Middle East and North Africa region for the period 2010–2018 using both the three-stage least-square and the generalised method of moments. The results provide strong evidence that banks located in countries with high exposure to terrorist attack exhibit low financial stability, due to high bank risk (i.e., high credit and insolvency risk). However, these banks show high financial performance (i.e., high profitability and cost efficiency), on average. Our results also show differential impacts on bank stability for countries marked as more (less) exposed to risk of attacks. For banks located in high-income-generating countries, we find that exposure to terrorism is associated with low financial performance and high credit risk, which is the opposite case for low-income-generating countries. Our results also indicate high systemic risk for listed banks operating under high terrorism risk exposure.
... In these 18 years, countries in the Middle East and North Africa (MENA) region have witnessed the highest level of terrorism due to various armed conflicts. MENA region is the cradle of many notorious terrorist groups like Islamic State (IS), Al-Qaida in the Arabian Peninsula (AQAP), etc., which are responsible for many devastating domestic and transnational terror attacks [2]. For example, ISIS (Islamic State of Iraq and Syria), also known as ISIL (Islamic State of Iraq and the Levant), is a Sunni jihadist group with extreme violent ideology. ...
... Therefore, the focus was on terrorism as an immanent political practice, the purpose of which is to achieve certain goals. Consequently, supporting terrorism required resources for reproduction (for example , Levitt 2002;Bjørgo 2005;Kim & Sandler 2020), and had consequences in the form of a geopolitical reaction (sanctions, military intervention, exclusion from the international system). Terrorism is an example of a choice between the internal political balance and the geopolitical ambitions of a particular political regime. ...
Full-text available
The article is devoted to the analysis of geohistorical transformations in the MENA region, which became the prerequisites for the Arab Spring. Author proposes to analyze the world-systemic retrospective of the Middle East, which is characterized by the clash of empires, the confrontation of religions and the geopolitical ambitions of many actors. Author defines three directions in the analysis of the prerequisites for the Arab Spring: (1) the region's imperial past and post-imperial transformations, (2) successful and unsuccessful modernization of political systems, (3) configurations of sources of social power. I assume that after the collapse of an empire, there always appears a political vacuum. In such vacuum, the political system either remains unchanged or completely changes its institutional design. In the case of the MENA region, many states that gained independence in the 20 th century reproduced the imperial logic of government, which consisted in extrapolating the center-peripheral model of political power to their own political systems. In addition, the dominance of the military type of power, as a continuation of the imperial logic of administering the periphery, together with nationalism, secularism, and geopolitical ambitions, played a key role in the emergence of bifurcation processes. Moreover, the unsuccessful unilateral modernization of independent states has weakened the institutional capacity for the stability of political systems. On the other hand, there is the example of Turkey, Iran, and the Gulf countries, which demonstrates successful modernization, the creation of a strong ideological power (with an emphasis on either nationalism or Islamism), the imperial past as centers of imperial worlds, the consolidation of various types of power. These processes led to the creation of not just modern states, but autonomous political systems capable of adapting to geopolitical configurations, leveling bifurcation processes, skillfully managing the mechanisms of control over the territory, population and security of the political system.
This study investigates whether women directors’ attributes affect market valuation for banks. Functional attributes such as independence and leadership are considered together with professional attributes such as financial expertise, education and nationality. We have constructed a unique sample of 1,019 bank-year observations for the period 2007-2017 for 12 developing countries that are characterized by low women’s empowerment/quotas and a dual banking system. Alternative measures for women on the board have been used (i.e. the percentage of women and a dummy indicator). Our findings suggest strong evidence that the presence of women directors on the board is positively associated with bank value. Women as independent board members are also positively associated with market value, whereas women being in a chairperson leadership role has no significant association. Accounting and finance qualifications are positively associated with bank market value, whilst conversely, women directors with a high level of education and those holding international qualifications or whom are foreign, are negatively associated with bank market value. As a mediating effect, we examine the cultural value orientations (i.e. cultural openness to diversity) for our sample countries. Our results demonstrate that women directors have a positive association with bank value in countries which are more open to diversity. The findings regarding women directors’ attributes tend to vary depending on the level of culture openness. We additionally examine the impact of different bank types (i.e. Islamic versus conventional banks) and the financial crisis of 2007.
Full-text available
While a considerable number of studies examine the nexus between terrorism and economic growth, there appears to be a dearth of research on the relationship between terrorism and public debt. The main objective of this chapter is to study the impact of terrorist attacks on the debt-to-GDP and sovereign credit rating for a panel of 11 MENA countries over the period 2002–2018. Given the presence of cross-sectional dependence and the heterogeneity of the panel, we apply the second-generation econometrics tests to investigate the long-run relationship between the variables. Regression techniques were based on the fully modified ordinary least-squares (FMOLS) estimation and Dumitrescu and Hurlin (Econ Model, 29(4):1450–1460, 2012) panel Granger causality test. Our findings highlight the presence of significant long-term relationship between the terrorism index and the debt ratio on the one hand and terrorism index and sovereign credit rating on the other hand. We notice that the violent attacks affect the credit rating more than the debt ratio. On average, one-point increase in terrorism index generates approximately four-notch downgrading for target country. Based on these results, several policy recommendations for the MENA countries are proposed.
This paper quantifies how past transnational terrorist attacks against a potential donor's assets result in enhanced foreign aid flows to a country hosting the responsible terrorist group. Given the reversed causality between foreign aid and terrorism, our empirical analysis puts forward an instrumental variable. Both conflict and governance assistance are shown to stem from transnational terrorist incidents involving recipient–donor dyads during 1974–2013 for a global sample. For recipient-related terrorism, lagged transnational terrorist events against a donor's assets display a robust positive impact on conflict and governance aid. Placebo or falsification tests support the exogeneity of the instrument.
This study investigates the effects of terrorist attacks on foreign investment by stressing the importance of both business- and non-business-targeting terrorism in the host country. Building on North’s strand of institutional theory, we argue that both forms of terrorism represent exogenous risks likely to generate high levels of non-ergodic uncertainty for MNEs and subsequently deter foreign investors. Further, we hypothesize that these effects may be moderated by host-country political regime type, which serves as a gauge for a favorable investment environment for MNEs operating in institutionally fragile markets. Using panel data on fifteen MENA countries over the period 2001–2018, we find empirical support for our hypotheses whereby hybrid political regimes, namely anocracies, strengthen the negative effects of both business- and non-business-targeting terrorism on FDI. Our work contributes to the research on FDI and exogenous risks by offering a more fine-grained conceptualization of terrorism, as well as by highlighting the moderating role of host-country hybrid regimes.
Drawing from statements by politicians, the media, policy analysts, and researchers, the current study identifies nine myths associated with terrorism and the practice of counterterrorism. We focus on those myths that have special policy relevance since the four al‐Qaida hijackings on September 11, 2001, and the ensuing heightened security concern. Many of those myths generated research articles that, at times, come to contradictory conclusions. Our goal is to provide recent statistics and a literature evaluation to sort out such contradictory results. In the case of the alleged macroeconomic consequences of terrorism to a typical country, we supply updated estimations. Throughout our presentation, we draw from the post‐2001 literature or statistical evidence. Often, empirical methods and procedures have evolved to a point where more clear‐cut and robust findings are now available through better identification and advanced estimation procedures.
The 9/11 attacks and the “War on Terror” brought terrorism and counterterrorism to the forefront of politics. Today, terrorism remains one of the most serious threats to national and international security. Yet, amid the surge of scholarly and policy interest in terrorism and counterterrorism, the literature on counterterrorism policies and strategies in the Middle East at the local and regional levels is sparse, limited and predominantly Western. The experience of political terrorism and violence in the Middle East throughout the 1980s and 1990s, which led to the introduction of a securitization process and the adoption of counterterrorism measures in the region long before the 9/11 attacks, offers important lessons. This paper takes its point of departure from the definitional conundrum of the concepts of “terrorism” and “(national) security” within the Middle East context. It examines the evolution of the securitization process in the Middle East in response to terrorism, with reference to the experiences of countries in the region. As it analyzes the different counterterrorism models and methods employed, it argues that counterterrorism strategies in much of the Middle East have not been effective and have at times been counterproductive.
Full-text available
This paper unpacks cyber radicalization and the rising threat of terrorism in the Middle East and North Africa region and Tunisia, mainly during the post-2011 Arab Spring Revolution, using semi-structured interviews based on qualitative research method and policy briefings analysis as a theoretical framework. The scope of the study will primarily focus on examining the key factors that led to the unprecedented rise of extremism in the MENA and Tunisia following the fall of significant dictatorships. Mainly, it seeks to build an inclusive understanding around the vulnerability model (Push and Pull Factors); that encompasses socio-economic grievances, identity crisis, and lack of sense of belonging besides state coercion and failure. Moreover, It will provide a brief background about the online strategic tactics used by terrorist organizations like ISIS, using examples of visuals and stats from Twitter and YouTube, etc. Additionally, the analysis of a few actors’ policy briefs fighting the lethal effect of online terrorism will help better evaluate the efficiency of international development initiatives in eliminating the destructive impact of such online content. Finally, this paper hopes to present an inclusive understanding of Tunisia’s struggle with extremism out of the colonial or white-centric frameworks that had contributed to estranging the country and its struggle.
Full-text available
This article introduces an extended data set of 760 terrorist groups that engaged in attacks during 1970 to 2016. Unlike most extant group data sets, the extended data on terrorist groups (EDTG) is not tied to terrorist groups and attacks listed in the RAND terrorism data; rather, EDTG is linked to terrorist groups and attacks given in the Global Terrorism Database. Terrorist groups’ variables in EDTG include ideology, main goals, start date, duration, base country, attack diversity, peak size, alternative endings (if relevant), and others. We display interesting features of EDTG through a series of tables and figures. Our EDTG-based survival analysis is at odds with some of the literature: for example, the demise of a leader and a larger share of transnational terrorist attacks increase the group’s odds of failure. After 2001, religious terrorist groups are more resilient than those with other ideologies. We also analyze terrorist group lethality and productivity.
This overview examines critically the post-9/11 empirical literature on terrorism. Major contributions by both economists and political scientists are included. We focus on five main themes: the changing nature of terrorism, the organization of terrorist groups, the effectiveness of counterterrorism policies, modern drivers or causes of terrorism, and the economic consequences of terrorism. In so doing, we investigate a host of questions that include: How do terrorist groups attract and retain members? What determines the survival of terrorist groups? Is poverty a root cause of terrorism? What counterterrorism measures work best? In the latter regard, we find that many counterterrorism policies have unintended negative consequences owing to attack transference and terrorist backlash. This suggests the need for novel policies such as service provision to counter some terrorist groups’ efforts to provide such services. Despite terrorists’ concerted efforts to damage targeted countries’ economies, the empirical literature shows that terrorism has had little or no effect on economic growth or GDP except in small terrorism-plagued countries. At the sectoral level, terrorism can adversely affect tourism and foreign direct investment, but these effects are rather transient and create transference of activities to other sectors, thus cushioning the consequences. (JEL F21, F52, H56, K42, Z31)
This article reports on trends in organized violence and peace agreements collected by the Uppsala Conflict Data Program (UCDP). The number of fatalities in organized violence decreased for the fourth consecutive year, to reach the lowest level since 2012. In 2018, UCDP recorded almost 76,000 deaths: a decrease of 20% compared to 2017, and 43% compared to the latest peak in 2014. State-based armed conflict drives this downward trend in organized violence, with Syria accounting for much of the change. The number of civilians killed in one-sided violence also dropped in 2018, reaching its lowest level since 2012. In contrast, non-state conflict remained on a high level. The general decline in fatalities from organized violence does not correspond with the trend in the number of active conflicts. In fact, the world has seen a new peak in the number of conflicts after 2014, matched only by the number of conflicts in the early 1990s. In 1991, the peak in the number of armed conflicts corresponded with a similar peak in the number of signed peace agreements. This was followed by a decrease in the number of conflicts in the late 1990s and early 2000s. However, the most recent rise in armed conflicts has not been matched by a similar rise in the number of peace agreements. Two circumstances that characterize the recent rise in conflicts have also been found to make conflicts harder to solve: explicit religious claims and high levels of internationalization.
Regime type has opposing effects on terrorism. If a regime constrains the executive branch, then terrorism may be more prevalent. If, however, a regime allows all viewpoints to be represented, then grievances may be held in check, resulting in less terrorism. Regimes that value constituents’ lives and property will also act to limit attacks. We formulate a game-theoretic model, containing a terrorist group and targeted government, that captures these opposing forces and supports a nonlinear relationship between regime type and terrorism. This model indicates how diverse samples in the literature can result in different relationships between regime type and terrorism. Seldom does it support the positive relationship that is prevalent in the terrorism literature. We apply a large variety of empirical techniques to show that regime type has a robust inverted U-shaped impact on various terrorism measures. Foreign policy variables (e.g., alliance with the United States) are not a robust influence on terrorism.
The article uses newly available consistent military expenditure data for 1960–2014 to examine past and current global spending trends during and after the Cold War. We are particularly interested in the impact of the end of the Cold War, 9/11 and the 2008 recession on military spending worldwide. The global share of military spending of East Asia & Pacific and the Middle East & North Africa increased relative to other regions since 1985. This increase underscores the need for western allies to bolster their power projection capacities. After 1999, both China and Russia raised their real defense spending, with China's increases far exceeding that of Russia. Both countries have a long ways to go to rival US capabilities. The North Atlantic Treaty Organization (NATO) continues to be heavily dependent on US defense spending. The post-1998 expansion allies assume about 2.4 per cent of the alliance's defense burden, while representing significant risks to NATO, given recent Russian actions in the Ukraine.
Students in both social and natural sciences often seek regression methods to explain the frequency of events, such as visits to a doctor, auto accidents, or new patents awarded. This book, now in its second edition, provides the most comprehensive and up-to-date account of models and methods to interpret such data. The authors combine theory and practice to make sophisticated methods of analysis accessible to researchers and practitioners working with widely different types of data and software in areas such as applied statistics, econometrics, marketing, operations research, actuarial studies, demography, biostatistics and quantitative social sciences. The new material includes new theoretical topics, an updated and expanded treatment of cross-section models, coverage of bootstrap-based and simulation-based inference, expanded treatment of time series, multivariate and panel data, expanded treatment of endogenous regressors, coverage of quantile count regression, and a new chapter on Bayesian methods.