Drivers and Persistence of Death in Conflicts: Global Evidence

Abstract

Spanish

Chinese

We investigate persistence and determinants of deaths from conflicts in a sample of 163 countries for the period 2010–2015. The empirical evidence is based on the Generalized Method of Moments. First, the findings are contingent on income levels, religious domination, landlockedness, regional proximity, and legal origins. We find that the persistence of deaths in internal conflict is more apparent in coastal, French civil law, and Islam-oriented countries, compared to landlocked, English common law, Christian-oriented countries, respectively. Second, the following factors are generally responsible for driving deaths from internal conflicts: homicides, conflict intensity, and conflicts fought. Furthermore, incarcerations have negative effects on internal conflicts. Justifications for the established tendencies and policy implications are discussed.

Keywords

War Conflicts Global Evidence Persistence Violence Homicide Deaths from Conflicts Comparative International Studies Empirical Studies Conflict Resolution

Three main tendencies in scholarly circles constitute the premise of this article: the growing cost of global conflicts, the policy relevance of elucidating the causes of deaths from conflicts, and gaps in the literature. First, conflict-related negative externalities (including death) are increasingly a global concern. This is sustained by the 2015 Global Peace Index (GPI) report which maintains that about 13 percent of the annual wealth produced in the world is allocated to tackle conflict-related issues (Anderson 2015a; Asongu and Kodila-Tedika 2017). This annual wealth is equivalent to the Gross Domestic Product (GDP) of the following countries: Spain, France, Brazil, the United Kingdom, Canada, and Germany. Given these sober facts, the significant amount of wealth spent on managing deaths from conflicts could otherwise have been used to fund socioeconomic projects related to sustainable development goals. This narrative is broadly consistent with the literature on what it takes to resolve conflicts and promote peace (Bodine and Crawford 1998; Lenski 2014; Malhotra 2016; Rubenstein 2017).

Second, building on the above, it is policy-relevant to understand the drivers of internal and external conflicts in the world. This is important given that they could be leveraged upon to potentially prevent conflicts and the corresponding costs associated with conflict-related deaths. Mitigating, avoiding, and preventing such “deaths in conflicts” would also logically have positive socioeconomic externalities on society as a whole.

Third, from a scholarly viewpoint, this study is relevant because of the sparse literature on drivers of deaths from conflicts around the world. To the best of our knowledge, while there are studies on trends and patterns in battle deaths, the extant literature on drivers of deaths in conflicts around the world has not focused on persistence using more contemporary data. For instance, Lacina and Gleditsch (2005) engaged patterns and trends in deaths from battles using a (then) new data set made available to the scientific community. Murray and others (2002) focused on deaths from conflicts within the framework that conflicts are understood as an issue of public health. Obermeyer, Murray, and Gakidou’s (2008) “50 years of violent war deaths” from Vietnam to Bosnia builds on a data analysis from the world health survey program. Lacina (2006) viewed the “severity of civil wars” in light of the number of deaths registered in civil wars, while Lacina, Gleditsch, and Russett (2006) argued for a “declining risk of death in battle.” The impact of conflicts on adult mortality using a time series cross-national analysis was examined by Li and Wen (2005).

To increase options for policy implications, the empirical analysis offered here emphasizes a multitude of comparative fundamentals. To this end, the data set is decomposed into fundamental characteristics contingent on legal origins, income levels, religious domination, landlockedness, and regional proximity.¹ The rest of the article is organized as follows. Next we offer an overview noting some of the most important conflicts in the world today by region. We then discuss the theoretical underpinnings and related literature, after which the data and methodology are presented. We next cover empirical results and the discussion, before concluding with implications and future research directions.

Regional Overview

While by no means exhaustive, the following notes highlight very broadly by region, the major conflicts—both internal and external, war-related and otherwise—across the world.

Sub-Saharan Africa (SSA)

SSA is experiencing substantial deaths from conflicts, partly due to increased levels of terrorism and partly because of its contemporary history of political strife and civil wars (Kossele 2020; Kossele and Shan 2018). First, among the terrorist organizations inflicting casualties in SSA, the Boko Haram of Nigeria is the most prolific and has extended its activities to neighboring Cameroon, Chad, and Niger. In 2014, Boko Haram was responsible for 6,664 deaths compared to 6,074 from Islamic State of Iraq and the Levant (ISIL; Searcey and Santora 2015). Complementary insights that could be driving the persistence of deaths from conflicts in SSA include South Sudan’s civil war between the Dinkas and Nuers, which in 2014 caused 1,817 deaths, and al-Shabaab, which in the same year was responsible for 4,425 deaths, including 633 civilians (Anderson 2015b; Asongu et al. 2018a, 2018b).

Second, in the contemporary era SSA has been characterized by more political instability and violence compared to other regions of the world. Although some of these conflicts are not in the sampled periodicity (2010–2015), there are recurrent pockets of political unrest across the region. Indeed, “seven of the nine cases of total chaos and societal breakdowns known in recent history have been registered in Africa (with the exceptions of Afghanistan and Syria): Angola, Burundi, Sierra Leone, Liberia, Zaire/Congo, Somalia, and Sudan” (Asongu 2014b, 1569). Recently, the internal conflicts in South Sudan (Justin and de Vries 2019; Krause 2019) and the Western part of Cameroon (Keke 2020; Mbih 2020; Nganji and Cockburn 2020) are similarly indicative of the fact that internal conflicts remain a critical setback to economic development in SSA.

Middle East and North Africa (MENA)

Conflicts in the MENA region include, inter alia, the Israeli-Palestinian conflict; concerns surrounding the Iranian nuclear weapons program; the Iraqi political stalemate; the proxy war in Yemen; and the Syrian conflict.

The Israeli–Palestinian conflict remains one of the oldest conflicts that have consistently resulted in deaths over the past 70 years (Ben-Meir 2020). The issue is fundamentally based on the inability of either country to acknowledge the rights of the other to the same piece of land. This has led to endless violence between Palestinians and Israelis and resolving the conflict would require courageous and visionary leaders prepared to make concessions for the peaceful co-existence of both countries on the same piece of land.

Concerns surrounding the Iranian nuclear weapons program began in the aftermath of the 1979 Iranian revolution, an era during which Iran sought to lead the region by becoming a nuclear power (Ben-Meir 2020). The ramifications of Iran’s quest to become a regional superpower include the complicity of the country in conflicts that affect political instability in the neighboring countries of Syria, Iraq, Yemen, and Lebanon. The United States recently withdrew from a deal meant to contain Iran’s nuclear weapons agenda and this has encouraged Iran to continue on her path to becoming a nuclear power in the region (Laipson 2019; Pieper 2019), despite crippling U.S. sanctions. Moreover, the recent assassination of Iranian General Soleimani in Iraq by the United States has increased tensions in the region as many are concerned that the U.S.–Iranian strife could escalate (Singh 2020).

The Iraq war of 2003 continues to have repercussions on peace and stability in the Middle East today. Even prior to 2003, Iraq has experienced many waves of political strife and turbulence since its establishment in 1932—notably the 1958 revolution, Saddam Hussein’s rise to power in 1979 which engendered the Gulf and Iran–Iraq wars, and the 2003 war that has led to more than 100,000 deaths and substantially damaged infrastructure (Ben-Meir 2020). In addition, cross-border conflicts and requests for more autonomy by some regions of the country have led to recurrent conflicts in the contemporary era (Akdedian and Hasan 2020; Zarei 2020).

The Yemeni war is considered by most analysts as one of the most horrific human disasters in modern history, not least because it is a proxy war involving two countries also fighting to become regional superpowers: Iran and Saudi Arabia (Ben-Meir 2020). While Iran is backing the leader of Shi’ite Muslims through Houthi rebels, Saudi Arabia is supporting an internationally recognized government led by Sunni Muslims (Darwich 2020). The war has led to over one million children sick of cholera, thousands of deaths, and millions left to starve, with many dying of starvation (Ben-Meir 2020; Blackburn, Lenze, and Casey 2020).

Despite the fact the Syrian government now has an upper hand in the Syrian conflict, the underlying conflict is one of the deadly externalities of the Arab Spring. It has had devastating consequences in terms of the number of refugees and internally displaced persons (over 11 million) and the number of people killed as a result of the war (about 700,000) (Ben-Meir 2020; David et al. 2020). Despite some evidence that a proxy war is also being fought in the country between the United States, Turkey, Russia, and Iran, it is apparent that President Assad is increasingly restoring his authority across the country (Walther and Perdersen 2020). However, until external forces leave and the rebels surrender, the conflict is unlikely to stop abruptly.

Latin America

In contrast, the persistence of deaths in Latin America is much more related to internal conflicts, particularly the high rate of homicides in the region. According to Muggah and de Carvalho (2017), one youth in Latin America is murdered every 15 minutes. Seven countries in this region account for approximately one-quarter of annual homicides worldwide: Venezuela, Mexico, Honduras, Guatemala, El Salvador, Colombia, and Brazil (see Luhnow 2018). In 2016, Latin America was host to 43 of the 50 most murderous cities in the world (The Economist 2017), and that number has dropped by only one city in 2020.²

Europe

Ukraine has witnessed the most notable contemporary crisis in Europe. It has seen a power struggle between two factions in the country, one aligned with Russia, the other supported by the European Union (Amadeo and Boyle 2020). Ukraine was a crucial contributor to the economy of the Soviet Union between 1920 and 1991. Yet, following a political crisis from 2014 to 2018, a military conflict ensued between Russian-backed separatists and Ukrainian soldiers.

A possible reason for anticipating low persistence of death from European conflicts involves its comparatively high capacity to resolve concerns surrounding internal conflicts. For instance, while only 20 percent of murders are solved in the Latin American continent, approximately 80 percent of homicides are resolved in European countries (Muggah and de Carvalho 2017). This comparative ability to address conflict scenarios may be fundamentally traceable to the peoples’ belief in government institutions as a way to resolve conflicts. By contrast, where government institutions fail to engender sufficient trust, people often resort to conflicts as means of making their voices heard in the absence of a stringent rule of law (Choi 2010) and good governance (Asongu and Nwachukwu 2017d).

Asia

Asia has endured several extremely bloody crises and conflicts, most notably in Myanmar, Kashmir, the Tamil crisis in Sri Lanka, and the Sikh Minority in Pakistan. Perhaps the most notable contemporary crisis over the past decade has involved the crackdown of the Rohingya Muslims by Myanmar’s army leading to hundreds of thousands fleeing to Bangladesh. The crackdown, which began in August 2017, has been considered by the United Nations (UN) as a “textbook example of ethnic cleansing” (BBC News 2020; Druce 2020). As of January 2020, a top UN court ordered that measures should be taken by Myanmar’s Buddhist-majority to protect Rohingya community members from genocide. While some government measures are being taken, the country’s army is still denying allegations of genocide by arguing that they are targeting Rohingya militants and not civilians (Islam 2020).

Theoretical Underpinnings and Literature Review

The theoretical framework of persistence underpins much of the literature in inclusive development (Asongu and Nwachukwu 2017a) and financial development (Goddard et al. 2011; Stephan and Tsapin 2008). The approach is also important to studies focusing on income-level convergence which constitute a substantial part of neoclassical growth models (Barro 1991; Barro and Sala-i-Martin 1992, 1995; Baumol 1986; Mankiw, Romer, and Weil 1992). These theoretical insights are increasingly being transplanted to other areas such as the development of stock markets (Asongu 2013; Bruno, de Bonis, and Silvestrini 2012; Narayan, Mishra, and Narayan 2011) and non-exclusive human progress (Asongu 2014a; Mayer-Foulkes 2010).

Novel theories of economic growth became relevant principally as a result of improvements in the neoclassical revolution that led to a considerable shift in income differences across nations. Under this theoretical canopy, concepts of market equilibrium were developed to emphasize the importance of theories of economic growth which predicted absolute decreases in income levels across countries. Consistent with Mayer-Foulkes (2010), successful convergence trends are fundamentally traceable to positive externalities from “free market competition.” Two main strands dominate the attendant literature. One argues that only divergence (i.e., lack of convergence) is feasible because of, inter alia, disparities in initial conditions, the presence of multiple equilibria, and multiple endowments (Barro 1991; Pritchett 1997). Another strand claims that, despite heterogeneities in initial conditions, cross-country variations in levels of income are nevertheless feasible, especially when countries converge to a steady state or long-run equilibria (Asongu and Nwachukwu 2017a). The purpose of this study is not to take positions for or against the contending schools of thought. What is relevant here is a common denominator: the criterion used to assess and draw conclusions on the presence or absence of convergence. The common criterion applied in this study is that for convergence to be established, the estimated lagged dependent variable should be within the interval of zero and one.

Table 1 offers a useful summary of key literature on factors driving or deterring conflicts across the world. To the best of our knowledge, while there are studies on patterns and trends in deaths from battles (noted earlier), scholarship on the persistence and drivers of death in conflicts using more contemporary global data is sparse. Hence, the added value of this study is to complement the existing literature by using more contemporary global data and focusing on deaths as a specific outcome of conflicts, that is, internal and external deaths.

Table 1.

Factors Affecting Conflicts.

Author(s)	Period	Countries/others	Methodology	Conflicts	Instruments	Results
Costa, Hermandez, and Sebastian-Gallés (2008)	Undisclosed	Two groups of 100 participants	Error analysis	Conflicting information	Bilingualism	Resolves conflicting information
Gubler and Selway (2012)	1945–1999	159 countries	Logit model	Civil wars	Societies where ethnicity is cross-cut by socioeconomic class, geographic region and religion	Civil war onset is an average of nearly 12 times less probable
Jo and Namgung (2012)	1957–2008	221 PTAs	Probit analysis	Conflicts in PTAs	DSMs	Democracies are more likely to engage DSMs
Akcinaroglu and Radziszewski (2013)	1990–2008	African countries with civil wars	Cox proportional hazards model	Civil wars in Africa	PMCs	Duration of conflict reduces with PMCs are in competition
Brancati and Snyder (2013)	1945–2008	Post-World War II civil war cases	Matching methods and logistic regression	World civil wars	Hold post-war elections after establishing strong institutions.	Quick post-war elections fuel post-war conflicts
Adhikari, Hansen, and Powers (2012)	1996–2006	Nepal	Probit analysis	Civil war	Understanding loss and risk factors	Redress of war grievances
Walter (2015)	1945–2009	Oslo (UCDP/PRIO) Armed Conflict Database (ACD) Version 4.0 (181 countries)	Cox Proportional Hazards Models—Reduced Models and Fixed effects models	Inter-state conflicts	Political institutions	Strong political institutions mitigate war in post-conflict societies.
Asongu and Nwachukwu (2016b)	1996–2006	MENA countries	GMM System	Political and socioeconomic chaos	Catch-up in negative signals like inflation, unemployment, and bad governance	From 2007, the Arab Spring could occur between 2011 and 2012
Asongu and Nwachukwu (2016e)	1996–2010	53 African countries	GMM System	Political stability and non-violence	Education and lifelong learning	Instruments favor political stability
Asongu and Kodila-Tedika (2016)	2009–2010	38 African countries	OLS	Conflicts and crimes	Six good governance tools	Corruption-control is the best tool

Note. PTA = Preferential trade agreement; DSM = Dispute settlement mechanism; PMCs = Private military companies; UCDP = Uppsala Conflict Data Program; PRIO=Peace Research Institute Oslo; MENA = Middle East and North Africa; GMM = Generalized methods of moments; OLS = Ordinary least squares.

We now briefly discuss the studies summarized in Table 1. Building on the curiosity that the imperative of bilinguals to continuously employ two languages during a speech could affect their networks of attention, Costa, Hermandez, Sebastian-Gallés (2008) compared the performance of monolinguals and bilinguals in a previously developed attentional network task. They show that participants with a bilingual culture were faster than monolinguals at task performance and more efficient when it came to executive alerting and executive control networks. Bilinguals were not only better at resolving conflicting information, they could also leverage more on an alerting cue. In addition, reduced switching cost between a multitude of trials was experienced by bilinguals compared with monolinguals. Bilingualism is also established to exert some influence on young adults’ abilities to attain efficient attentional channels.

Using a data set on preferential trade agreements (PTAs) between 1957 and 2008 to assess what factors determine the legal arrangements of dispute settlement mechanisms (DSMs), Jo and Namgung (2012) established three main findings: (1) it is more likely for democracies to prefer DSMs that are moderately strict, when compared with autocracies; (2) similar templates are increasingly adopted by each trading partner in their attempt to emulate each other; and (3) recent evidence of legalistic channels substantially build on the development of trade regimes that are multilateral. By highlighting the relevance of both macro-level and member-specific factors, the results have valuable implications for the conception of international institutions drawn on in this study and others.

Gubler and Selway (2012) present a comprehensive theory of civil war onset based on ethnic groups, geographic dispersion, and horizontal inequality. The mobilization efforts of rebel leaders are thwarted in societies with highly cross-cutting tendencies. This is essentially because of a lower possibility of potential combatants to identify with nationalist ambitions, reduced communication in groups, and limited ability to manifest social control. The authors conclude that, on average, the onset of a civil war is about twelve times less likely in environments where ethnicity is influenced by religion, geographic region, and socioeconomic class.

Adhikari, Hansen, and Powers (2012) used data in Nepal to assess the demand for post-conflict reparations by individuals. They concluded that the understanding of risk and loss features is imperative in civil reconciliation and settlement. Brancati and Snyder (2013) argued that when elections are held immediately following a civil war, it can increase the possibility of renewed fighting. They concluded that conditions that can mitigate risks include peacekeeping, demobilization, power sharing, decisive victories, and strong institutions. Asongu and Kodila-Tedika (2016) analyze the role of governance tools (corruption-control, rule of law, regulation quality, voice/accountability, government effectiveness, political stability/non-violence) in fighting conflicts and crimes to conclude that corruption-control is the most effective. This is broadly consistent with Walter (2015) who concluded that bad governance leads to a repeat of war and conflict trap.

The effect of African private military companies (PMCs) on the length of a conflict is investigated by Akcinaroglu and Radziszewski (2013) from 1990 to 2008. The empirical findings show that, with increasing competition among government-hired PMCs, optimal service delivery is enhanced and thus promotes the swift end of conflicts. Asongu and Nwachukwu (2016b) predict the occurrence of the Arab Spring in a sample of MENA countries, whereas Asongu and Nwachukwu (2016e) demonstrate that education and lifelong learning can be used to reduce political conflicts and violence.

Intuition for Comparative Development Fundamentals

This section substantiates the premise for disaggregating the sample in terms of legal origins, regions, religious dominations, landlockedness, and income levels. These fundamental features of comparative economic development have been documented in several comparative development studies (Asongu and le Roux 2017; Beegle et al. 2016; Mlachila, Tapsoba, and Tapsoba 2017; Narayan, Mishra, and Narayan 2011). The comparative features of each variable are discussed in more detail in the subsequent paragraphs.

First, Asongu and others (2020a) argued that income levels elucidate the persistence of weapons and that income levels explain cross-country political instability and political terror (Asongu et al. 2020b). If they are correct, a logical extension would be that low income levels can be expected to be linked with higher levels of deaths from conflicts. Given that these authors argue that the wealth of nations explains political stability, political terrorism, and persistence in weapons, it is logical for this study to also posit that income levels can explain deaths from conflicts. This, we argue, is because low-income countries are likely to be more associated with political instability and conflicts owing to their limited financial resources—resources clearly needed to prevent conflicts as well as their corresponding negative externalities, which include death from conflicts and political instability. Similar arguments have been employed in contemporary studies to justify why foreign aid flows from developed to developing countries to mitigate terrorism’s negative externalities that are more apparent in developing countries (Asongu and Kodila-Tedika 2017; Efobi, Asongu, and Beecroft 2015).

Second, as documented in Arvis, Marteau, and Raballand (2007) and Asongu and le Roux (2017), landlockedness is closely associated with institutional and socioeconomic costs which are more likely to spark political conflicts (Asongu et al. 2020b). This extension is based on the fact that political instability is among the attendant institutional costs. It follows that coastal countries should be linked with less persistence in deaths in conflicts, compared to landlocked countries.

Third, we identify a nexus between religion and deaths from conflicts. As argued by Asongu and Nwachukwu (2017c), given the liberal nature of Christian-oriented countries relative to their Islam-related counterparts (which are comparatively more conservative), Christian-related countries are likely to be more linked to civil unrest and political terror. This is because citizens in such countries are liberal in their rights to express their grievances. The argument is broadly supported by Li (2005) who has posited that democratic institutions do provide an enabling environment for aggrieved citizens to employ mechanisms (which may lead to political instability) as means to resolve conflicts. Conflicts naturally engender deaths which are captured by the outcome variables used in the study. Compared to less liberal nations (which are mostly Islam-oriented countries) that are largely characterized as stable autocracies, more deaths owing to political instability could be related to unstable democracies because they do less to curtail ex ante violence. Put differently, compared to failed democracies, deaths from conflicts could be less linked to stable autocracies because such stable autocracies are more able to prevent political instability (Asongu et al. 2020a, 2020b; Drakos and Gofas 2006; Eubank and Weinberg 1994; Lai 2007; Piazza 2007, 2008; Schmid 1992).

Fourth, building on the law and economic development literature (Agbor 2015; Beck, Demirgüç-Kunt, and Levine 2013; La Porta et al. 1998, 1999), legal origins influence political instability/violence and corresponding deaths from conflicts due to political and adaptability frameworks. French civil law countries are more likely to be associated with deaths from conflicts since the post-independence politico-economic structure left by their former colonial power entails the substantial presence of the former colonial power. For instance, in addition to France having a tight control on the currency used by most of her former colonies in Africa, as of 2014, compared to countries with English common law, French civil law countries had registered substantially more coup d’états with the help of France; most of the coup d’états subsequently led to conflicts and deaths from conflicts (Koutonin 2014).

Fifth, regarding income levels, high-income nations are less associated with persistence in deaths from conflicts because they are less linked to political instability, civil wars, political strife, and violence (Asongu et al. 2020a). Even in the event of political instability and terror, high-income countries have the financial resources to dampen unfavorable political externalities such as loss of human life (Asongu et al. 2020b; Gaibulloev and Sandler 2009). This aligns with the premise that high-income countries are associated with better institutions with which to mitigate risks that, inter alia, are linked to deaths from civil wars and political strife (Anyanwu and Erhijakpor 2014; Asongu 2012; Asongu and Nnanna 2019b; Efobi 2015; Pelizzo and Nwokora 2016, 2018; Pelizzo et al. 2016). Note that the argument of income levels and regions can be valid in respect of employing them as fundamental characteristics of comparative development because (1) in regions that mostly consist of high-income nations, not all corresponding countries are high-income nations; and (2) in regions that consist of mostly countries with low incomes, some higher-income countries are also apparent.

Data and Method

Data

This article investigates a sample of 163 countries for the period 2010–2015. The data were obtained from the Operations of Criminal Justice Systems (CTS), Institute for Economics and Peace (IEP), the Uppsala Conflict Data Program (UCDP) Battle-Related Deaths Dataset, the United Nations Committee on Contributions; the United Nations Office on Drugs and Crime (UNODC) Surveys on Crime Trends, and a Qualitative Assessment by the Economic Intelligence Unit (EIU) analysts’ estimates. The choice of the periodicity was motivated by the need to have findings with more updated policy implications.³

Two outcome variables are employed: deaths from external conflict and deaths from internal conflict. The independent variable of interest is the lagged dependent variable, while the control variables include security officers and police, homicides, incarcerations, conflict intensity, conflicts fought, and the United Nations Peace Keeping Force (UNPKF). The variables in the conditioning information set have been documented in the conflicts and political instability literature (Adhikari, Hansen, and Powers 2012; Akcinaroglu and Radziszewski 2013; Asongu and Kodila-Tedika 2016; Gubler and Selway 2012; Jo and Namgung 2012; Walter 2015).

In light of the motivation of increasing policy implications, the data set is disaggregated into fundamental features based on (1) regions (North America, Latin America, South Asia, East Asia and the Pacific, Europe and Central Asia, MENA, SSA); (2) landlockedness (Landlocked and Coastal); (3) religious domination (Christian with Catholic domination, Christian with Protestant inclination, Christian countries in which another Christian religion apart from Catholicism and Protestantism is dominant, Buddhist-oriented and Islam-dominated countries); and (4) legal origins (French civil law, English common law, German civil law, and Scandinavian civil law and Socialist countries). These fundamental features have been adopted in recent literature on comparative economic development (Asongu and le Roux 2017; Asongu and Nwachukwu 2017c; Beegle et al. 2016; D’Amico 2010; Mlachila, Tapsoba, and Tapsoba 2017; Narayan, Mishra, and Narayan 2011). We now discuss the information criteria used to select the highlighted fundamental features.

First, the basis for legal origins is obtained from La Porta, Lopez-De-Silanes, and Shleifer (2008, 289). The categorization by religious domination is from the World Fact Book (Central Intelligence Agency 2011). The World Bank’s income-level categorization is used to classify countries in terms of income levels.⁴ Landlocked countries are directly apparent from a world map. Appendix A provides the definitions of variables and sampled countries, while summary statistics are disclosed in Appendix B. Appendix C provides the correlation matrix.

Method

The choice of the estimation technique is consistent with the behavior of the dependent variables and characteristics of the data structure. Following recent literature on the persistence of macroeconomic variables, we adopt the Generalized Method of Moments (GMM) as estimation strategy (Asongu and Nwachukwu 2017a; Doyle 2017; Tchamyou 2019, 2020). Four main motivations underpin the selection of this empirical strategy. First, the number of agents is higher than the number of periods. Hence, the N(163) > T(6) premise is satisfied because T represents the number of years, while N denotes the number of countries. Second, given that the estimation approach is panel-oriented, cross-country variations are considered in the estimations. Third, inherent biases in the difference estimator are corrected with the system estimator. Fourth, the technique accounts for endogeneity by employing instruments to address the concern of simultaneity. The control for time invariant variables also enhances the control for endogeneity because it accounts for the unobserved heterogeneity.

This study adopts the Roodman (2009a, 2009b) extension of Arellano and Bover (1995) which has been established to better control for cross-sectional dependence and to reduce over-identification and hence is more efficient compared to traditional difference and system GMM approaches (Asongu and Nwachukwu 2016c; Baltagi 2008; Boateng et al. 2018; Love and Zicchino 2006; Tchamyou, Erreygers, and Cassimon 2019).

The following equations in level (1) and first difference (2) summarize the standard system GMM estimation procedure.

W_{i, t} = σ_{0} + σ_{1} W_{i, t - τ} + \sum_{h = 1}^{6} δ_{h} X_{h, i, t - τ} + η_{i} + ξ_{t} + ε_{i, t}

(1)

\begin{array}{l} W_{i, t} - W_{i, t - τ} = σ_{1} (W_{i, t - τ} - W_{i, t - 2 τ}) + \sum_{h = 1}^{6} δ_{h} (X_{h, i, t - τ} - X_{h, i, t - 2 τ}) \\ + (ξ_{t} - ξ_{t - τ}) + (ε_{i, t} - ε_{i, t - τ}) \end{array}

(2)

where $W_{i, t}$ is an indicator of the conflicts (internal or external), in country i at period t, $σ_{0}$ is a constant, X is the vector of control variables (security officers and police; homicides, incarcerations, conflict intensity, conflicts fought, and the UNPKP), τ represents the coefficient of auto-regression which is one for the specification, $ξ_{t}$ is the time-specific constant, $η_{i}$ is the country-specific effect and $ε_{i, t}$ the error term.

In accordance with the literature, it is important to articulate the process of corresponding exclusion restrictions since these are essential for a robust GMM specification. As far as exclusion restrictions are concerned, all explanatory variables are considered as suspected endogenous or predetermined variables and exclusively time-invariant omitted indicators are considered as strictly exogenous variables. This identification approach is consistent with recent literature (Asongu and Nwachukwu 2016d; Boateng et al. 2018). This process has been substantiated by Roodman (2009b) who argued that it is very unfeasible for years or time invariant omitted variables to be first-differenced endogenous.⁵

With regard to exclusion restrictions, consistent with the identification process, the outcome variables are influenced by the strictly exogenous indicators exclusively via the suggested endogenous or predetermined channels, which are explaining variables. For the suggested exclusion restrictions to hold, the alternative hypothesis underpinning such a restriction should be rejected. Note that the alternative hypothesis to the Difference in Hansen Test (DHT) argues for the position that the time-invariant variables are not strictly exogenous. In other words, the null hypothesis argues that invariant indicators affect conflicts exclusively through the endogenous explaining channels.

Given the above clarifications, in the findings presented below, the assumption of exclusion restriction holds if the DHT linked to the time-invariant instrumental variables (IV) (i.e., years, eq(diff)) is not rejected. Such a criterion for validating exclusion restrictions is not different from the standard IV procedure which requires a rejection of the alternative hypothesis of the Sargan Overidentifying Restrictions (OIR) test in order for the engaged instruments to affect the dependent variable exclusively via the suggested channels or endogenous explaining mechanisms (Asongu and Nwachukwu 2016a; Beck, Demirgüç-Kunt, and Levine 2003).

Empirical Results and Preliminary Discussion

Our empirical findings are presented in Tables 2 to 5. Tables 2 and 3 show findings on deaths from internal conflicts. Deaths from external conflicts are disclosed later in Tables 4 and 5. The first table for each dependent variable category discloses findings corresponding to income levels, religious domination, and landlockedness. The second table discloses estimated results corresponding to regions and legal origins. The last column of all the tables shows results of the full sample. Four main information criteria are used to examine the validity of the GMM models.⁶ Based on these criteria, the models of the first two tables pertaining to death from internal conflict are overwhelmingly valid. However, the corresponding models focusing on death from external conflicts are not valid, for the most part, because of the post-estimation presence of residuals in the estimations. For this reason, we exclusively discuss the valid findings in Tables 2 and 3 in this and the subsequent section. We have chosen to present the results of Tables 4 and 5 at the end of this article to avoid publication bias, especially the bias of exclusively disclosing positive, significant, or strong results in scholarly circles (Asongu and Nwachukwu 2017b; Franco, Malhotra, and Simonovits 1991; Rosenberg 2005).

Table 2.

Persistence in Internal Conflict with Income Levels, Religious Domination and Landlockedness.

	Dependent variable: Internal conflict
	Income levels				Religious domination					Openness to sea		Full sample
	HI	UMI	LMI	LI	CC	CP	CO	Islam	Bhu	LL	NLL	Full sample
Constant	0.225***	−0.273	−0.928**	−0.561	−0.469***	1.140***	−0.412	−0.335	1.391	−1.224***	−0.139	−0.336
	(0.000)	(0.381)	(0.038)	(0.125)	(0.000)	(0.000)	(0.765)	(0.541)	(0.672)	(0.000)	(0.721)	(0.381)
Internal Conflict (−1)	0.774***	0.470***	0.639***	0.700***	0.703***	0.493***	0.689	0.709***	0.733	0.688***	0.775***	0.742***
	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.560)	(0.000)	(0.528)	(0.000)	(0.000)	(0.000)
Security Officers & Police	0.0002	−0.032	0.070	−0.001	0.041	−0.290***	0.375	0.157*	−0.274	−0.083	0.023	−0.040
	(0.586)	(0.455)	(0.238)	(0.982)	(0.231)	(0.000)	(0.130)	(0.052)	(0.688)	(0.212)	(0.752)	(0.663)
Homicides	0.0001	0.041	0.185**	−0.170	0.006	0.060**	−0.175	−0.239	−0.291	0.159***	−0.061	−0.016
	(0.900)	(0.463)	(0.013)	(0.163)	(0.774)	(0.027)	(0.708)	(0.217)	(0.718)	(0.004)	(0.624)	(0.885)
Incarcerations	−0.0007*	−0.111	−0.288***	0.493***	0.052	−0.181***	−0.243	−0.010	0.050	0.198***	−0.061	−0.089
	(0.089)	(0.195)	(0.000)	(0.001)	(0.157)	(0.006)	(0.317)	(0.867)	(0.870)	(0.000)	(0.484)	(0.334)
Conflict Intensity	−0.00006	0.278***	0.339***	0.089	0.025**	0.180***	0.541	0.231***	0.117	0.215***	0.205***	0.286***
	(0.903)	(0.000)	(0.000)	(0.263)	(0.049)	(0.000)	(0.126)	(0.000)	(0.792)	(0.004)	(0.000)	(0.000)
Conflict Fought	0.0001	0.617***	0.144***	0.172	0.421***	−0.143***	−0.030	0.091	−0.154	0.268***	0.077	0.226***
	(0.757)	(0.000)	(0.000)	(0.101)	(0.000)	(0.004)	(0.979)	(0.290)	(0.937)	(0.00)	(0.311)	(0.000)
UNPKF	−0.0001	−0.023	0.078**	0.015	0.004	−0.033	−0.051	0.056	0.015	−0.030*	0.054**	0.032
	(0.723)	(0.123)	(0.016)	(0.476)	(0.695)	(0.125)	(0.847)	(0.118)	(0.853)	(0.067)	(0.027)	(0.168)
AR(1)	(0.117)	(0.157)	(0.038)	(0.052)	(0.329)	(0.155)	(0.221)	(0.003)	(0.504)	(0.026)	(0.015)	(0.005)
AR(2)	(0.193)	(0.290)	(0.581)	(0.054)	(0.390)	(0.304)	(0.552)	(0.193)	(0.812)	(0.043)	(0.550)	(0.453)
Sargan OIR	(0.003)	(0.000)	(0.016)	(0.000)	(0.000)	(0.000)	(0.000)	(0.041)	(0.002)	(0.001)	(0.000)	(0.000)
Hansen OIR	(0.995)	(0.865)	(0.447)	(0.465)	(0.347)	(0.508)	(1.000)	(0.660)	(1.000)	(0.359)	(0.737)	(0.479)
DHT for instruments
(a) Instruments in levels H excluding group	(0.664)	(0.600)	(0.631)	(0.118)	(0.588)	(0.895)	(0.986)	(0.323)	(0.972)	(0.566)	(0.598)	(0.348)
Dif (null, H = exogenous)	(0.999)	(0.851)	(0.319)	(0.794)	(0.241)	(0.251)	(1.000)	(0.774)	(1.000)	(0.262)	(0.679)	(0.532)
(b) IV (years, eq(diff)) H excluding group	(0.975)	(0.680)	(0.867)	(0.746)	(0.617)	(0.797)	(1.000)	(0.728)	(1.000)	(0.353)	(0.707)	(0.399)
Dif (null, H = exogenous)	(0.973)	(0.976)	(0.041)	(0.099)	(0.091)	(0.097)	(1.000)	(0.328)	(0.986)	(0.385)	(0.534)	(0.573)
Fisher	24,488***	5,012***	291.55***	46.84***	12,405***	772.5***	31.66***	86.69***	47.37***	3,007***	44.51***	47.95***
Instruments	31	31	31	31	31	31	31	31	31	31	31	31
Countries	43	36	46	38	54	26	14	49	13	34	129	163
Observations	215	180	230	190	270	130	70	245	65	170	645	815

Note. The significance of bold values is twofold. (1) The significance of estimated coefficients and the Fisher statistics. (2) The failure to reject the null hypotheses of (a) no autocorrelation in the AR(1) and AR(2) tests and (b) the validity of the instruments in the Sargan and Hansen OIR tests. AR = Autocorrelation; HI = High-income countries; UMI = Upper-middle-income countries; LMI = Lower-middle-income countries; LI = Low-income countries; CC = Christian countries with Catholic domination; CP = Christian countries with Protestant domination; CO = Christian countries in which another Christian religion apart from Catholicism and Protestantism is dominant; Islam = Islam-dominated countries; Bhu = Buddhism-dominated countries; LL = Landlocked countries; NLL = Not landlocked countries; UNPKF = United Nations Peace Keeping Force; OIR = Over-identifying Restrictions Test; DHT = Difference in Hansen Test for Exogeneity of Instruments Subsets; Dif = Difference.

, **, ***: significance levels at 10%, 5%, and 1%, respectively.

Table 3.

Persistence in Internal Conflict with Regions and Legal Origin Dynamics.

	Dependent variable: Internal conflict
	Regions							Legal origins					Full sample
	SA	ECA	EAP	MENA	SSA	LA	NA	Eng.	Frch.	Ger.	Scand.	Social.	Full sample
Constant	na	0.552**	0.035	−1.316	−0.311	−0.204		−1.235***	−0.619	0.488***			−0.336
		(0.038)	(0.887)	(0.174)	(0.407)	(0.639)		(0.001)	(0.148)	(0.001)			(0.381)
Internal Conflict (−1)		0.453***	0.596***	0.808***	0.719***	0.680***		0.668***	0.736***	1.512***			0.742***
		(0.000)	(0.000)	(0.000)	(0.000)	(0.000)		(0.000)	(0.000)	(0.000)			(0.000)
Security Officers & Police		0.044	0.0006	−0.112	−0.025	0.052		0.132**	−0.121	−0.0005			−0.040
		(0.260)	(0.993)	(0.558)	(0.703)	(0.158)		(0.023)	(0.135)	(0.934)			(0.663)
Homicides		−0.102***	−0.004	0.348	−0.117	0.028		0.118*	0.121	0.0004			−0.016
		(0.000)	(0.937)	(0.186)	(0.324)	(0.613)		(0.078)	(0.341)	(0.947)			(0.885)
Incarcerations		−0.039**	−0.017	0.177	−0.028	−0.009		0.084	−0.093	0.0003			−0.089
		(0.033)	(0.788)	(0.396)	(0.884)	(0.885)		(0.256)	(0.331)	(0.951)			(0.334)
Conflict Intensity		0.201***	−0.034	0.395***	0.301***	−0.072		0.315***	0.331***	−0.0004			0.286***
		(0.000)	(0.509)	(0.003)	(0.006)	(0.286)		(0.000)	(0.000)	(0.916)			(0.000)
Conflict Fought		−0.032	0.436**	−0.134	0.242***	0.401***		0.113**	0.240***	−1.000***			0.226***
		(0.899)	(0.021)	(0.282)	(0.005)	(0.001)		(0.032)	(0.001)	(0.000)			(0.000)
UNPKF		−0.072***	0.004	0.010	0.026	0.011		−0.012	0.031	0.0002			0.032
		(0.000)	(0.651)	(0.871)	(0.314)	(0.673)		(0.566)	(0.326)	(0.916)			(0.168)
AR(1)		(0.016)	(0.121)	(0.065)	(0.148)	(0.409)		(0.108)	(0.031)	(0.338)			(0.005)
AR(2)		(0.217)	(0.103)	(0.179)	(0.504)	(0.276)		(0.243)	(0.756)	(0.336)			(0.453)
Sargan OIR		(0.000)	(0.000)	(0.074)	(0.001)	(0.000)		(0.000)	(0.005)	(0.000)			(0.000)
Hansen OIR		(0.609)	(0.995)	(0.993)	(0.541)	(0.998)		(0.157)	(0.221)	(1.000)			(0.479)
DHT for instruments
(a) Instruments in levels H excluding group		(0.201)	(0.482)	(0.469)	(0.894)	(0.957)		(0.797)	(0.241)	(0.981)			(0.348)
Dif (null, H = exogenous)		(0.832)	(1.000)	(1.000)	(0.280)	(0.987)		(0.055)	(0.282)	(1.000)			(0.532)
(b) IV (years, eq(diff)) H excluding group		(0.401)	(0.971)	(0.771)	(0.385)	(0.989)		(0.112)	(0.224)	(1.000)			(0.399)
Dif (null, H = exogenous)		(0.911)	(0.988)	(1.000)	(0.778)	(0.969)		(0.515)	(0.333)	(1.000)			(0.573)
Fisher		289.49***	437.15***	1146***	129.27***	4471***		212.66***	37.7***	1910***			47.95***
Instruments		31	31	31	31	31		31	31	30			31
Countries		48	18	20	44	23		50	87	20			163
Observations		240	90	100	220	115		250	435	100			815

Note. The significance of bold values is twofold. (1) The significance of estimated coefficients and the Fisher statistics. (2) The failure to reject the null hypotheses of (a) no autocorrelation in the AR(1) and AR(2) tests and (b) the validity of the instruments in the Sargan and Hansen OIR tests. AR = Autocorrelation; ECA = Europe and Central Asia; EAP = East Asia and the Pacific; MENA = Middle East and North Africa; SSA = sub-Saharan Africa; LA = Latin America; NA = North America; SA = South Asia; Eng. = English Common Law countries; Frch. = French Civil Law countries; Ger. = German Civil law countries; Scand. = Scandinavian Civil law countries; Social = Socialist countries; UNPKF = United Nations Peace Keeping Force; OIR = Over-identifying Restrictions Test; DHT = Difference in Hansen Test for Exogeneity of Instruments Subsets; Dif = Difference.

, **, ***: significance levels at 10%, 5%, and 1%, respectively.

Table 4.

Persistence in External Conflicts with Income Levels, Religious Domination and Landlockedness.

	Dependent Variable: External conflicts
	Income levels				Religious domination					Openness to sea		Full sample
	HI	UMI	LMI	LI	CC	CP	CO	Islam	Bhu	LL	NLL	Full sample
Constant	0.631***	1.297***	1.801***	0.887***	1.259***	0.761***	−1.388	0.890***	0.975	1.267***	1.598***	1.229***
	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.001)	(0.396)	(0.000)	(0.363)	(0.000)	(0.000)	(0.000)
External conflicts (−1)	0.217***	−0.061***	−0.140***	0.063***	0.016	−0.129***	1.169	−0.029	0.349	−0.079***	−0.071*	−0.061*
	(0.000)	(0.000)	(0.000)	(0.008)	(0.702)	(0.000)	(0.127)	(0.146)	(0.626)	(0.000)	(0.085)	(0.050)
Security Officers & Police	−0.026	−0.003	−0.049*	0.060	−0.055**	0.303***	0.063	0.010	−0.043	−0.0006	−0.075*	−0.051
	(0.197)	(0.867)	(0.098)	(0.258)	(0.026)	(0.003)	(0.564)	(0.800)	(0.889)	(0.988)	(0.068)	(0.112)
Homicides	0.009	0.016	−0.072	0.015	−0.024*	−0.035	0.192	0.083**	−0.0005	−0.037**	−0.021	−0.006
	(0.835)	(0.328)	(0.231)	(0.654)	(0.087)	(0.424)	(0.271)	(0.011)	(0.998)	(0.042)	(0.709)	(0.862)
Incarcerations	0.132**	−0.057***	−0.083	0.135**	0.008	−0.132	−0.178	−0.021*	−0.086	−0.092	−0.016	0.020
	(0.011)	(0.005)	(0.105)	(0.029)	(0.744)	(0.323)	(0.215)	(0.068)	(0.848)	(0.231)	(0.737)	(0.650)
Conflict Intensity	−0.010	−0.024***	−0.031	−0.115***	−0.023*	0.004	0.211	0.0002	−0.0004	0.039	−0.038	0.0007
	(0.707)	(0.001)	(0.200)	(0.003)	(0.058)	(0.962)	(0.207)	(0.993)	(0.997)	(0.143)	(0.239)	(0.979)
Conflict Fought	0.126**	−0.024**	0.008	0.003	0.013	0.012	−0.051	−0.027*	−0.152	0.012	−0.031	−0.012
	(0.012)	(0.026)	(0.684)	(0.880)	(0.448)	(0.801)	(0.785)	(0.065)	(0.557)	(0.505)	(0.497)	(0.653)
UNPKF	−0.028	−0.006	0.007	0.012	0.005	0.042***	0.243	−0.0001	0.103	0.031**	0.004	0.017
	(0.215)	(0.247)	(0.724)	(0.146)	(0.561)	(0.004)	(0.165)	(0.991)	(0.451)	(0.035)	(0.782)	(0.116)
AR(1)	(0.022)	(0.839)	(0.068)	(0.068)	(0.071)	(0.959)	(0.265)	(0.995)	(0.670)	(0.381)	(0.886)	(0.434)
AR(2)	(0.028)	(0.197)	(0.055)	(0.100)	(0.016)	(0.026)	(0.302)	(0.090)	(0.194)	(0.065)	(0.001)	(0.000)
Sargan OIR	(0.000)	(0.649)	(0.000)	(0.017)	(0.984)	(0.001)	(0.006)	(0.000)	(0.976)	(0.010)	(0.022)	(0.001)
Hansen OIR	(0.445)	(0.984)	(0.466)	(0.775)	(0.711)	(0.674)	(1.000)	(0.957)	(1.000)	(0.708)	(0.853)	(0.773)
DHT for instruments
(a) Instruments in levels H excluding group	(0.191)	(0.625)	(0.302)	(0.626)	(0.227)	(0.550)	(0.941)	(0.382)	(1.000)	(0.355)	(0.563)	(0.373)
Dif (null, H = exogenous)	(0.655)	(0.992)	(0.557)	(0.713)	(0.901)	(0.628)	(1.000)	(0.996)	(1.000)	(0.803)	(0.853)	(0.861)
(b) IV (years, eq(diff)) H excluding group	(0.360)	(0.961)	(0.535)	(0.737)	(0.588)	(0.569)	(1.000)	(0.950)	(1.000)	(0.654)	(0.807)	(0.780)
Dif (null, H = exogenous)	(0.583)	(0.842)	(0.290)	(0.572)	(0.737)	(0.679)	(1.000)	(0.630)	(1.000)	(0.573)	(0.643)	(0.466)
Fisher	508.68***	13.31***	8.77***	8.83***	2.63***	96.75***	18.05***	3.04***	4.85***	89.46***	4.82***	3.66***
Instruments	31	31	31	31	31	31	31	31	31	31	31	31
Countries	43	36	46	38	54	26	14	49	13	34	129	163
Observations	215	180	230	190	270	130	70	245	64	170	645	815

Note.The significance of bold values is twofold. (1) The significance of estimated coefficients and the Fisher statistics. (2) The failure to reject the null hypotheses of (a) no autocorrelation in the AR(1) and AR(2) tests and (b) the validity of the instruments in the Sargan and Hansen OIR tests. AR = Autocorrelation; HI = High-income countries; UMI = Upper-middle-income countries; LMI = Lower-middle-income countries; LI = Low-income countries; CC = Christian countries with Catholic domination; CP = Christian countries with Protestant domination; CO = Christian countries in which another Christian religion apart from Catholicism and Protestantism is dominant; Islam = Islam-dominated countries; Bhu = Buddhism-dominated countries; LL = Landlocked countries; NLL = Not landlocked countries; UNPKF = United Nations Peace Keeping Force; OIR = Over-identifying Restrictions Test; DHT = Difference in Hansen Test for Exogeneity of Instruments Subsets; Dif = Difference.

, **, ***: significance levels at 10%, 5%, and 1%, respectively.

Table 5.

Persistence in External Conflicts with Regions and Legal Origin Dynamics.

	Dependent variable: External conflicts
	Regions							Legal origins					Full sample
	SA	ECA	EAP	MENA	SSA	LA	NA	Eng.	Frch.	Ger.	Scand.	Social.	Full sample
Constant	na	0.534***	0.642**	1.136**	0.789***	1.046***	na	1.028**	1.321***	1.055***	na	na	1.229***
		(0.004)	(0.039)	(0.024)	(0.001)	(0.000)		(0.046)	(0.000)	(0.001)			(0.000)
External Conflicts (−1)		0.053*	0.453***	0.017	−0.059***	−0.001		−0.135***	0.006	0.011			−0.061*
		(0.066)	(0.007)	(0.901)	(0.000)	(0.969)		(0.000)	(0.870)	(0.902)			(0.050)
Security Officers & Police		−0.056***	−0.001	−0.063	0.063	−0.002		0.038	−0.059**	−0.033			−0.051
		(0.001)	(0.989)	(0.620)	(0.294)	(0.834)		(0.466)	(0.017)	(0.318)			(0.112)
Homicides		−0.004	0.014	0.070	−0.018	0.0003		0.050	−0.029	0.011			−0.006
		(0.839)	(0.940)	(0.830)	(0.645)	(0.987)		(0.468)	(0.400)	(0.851)			(0.862)
Incarcerations		0.0001	−0.015	0.045	0.118	0.0004		−0.027	−0.011	0.021			0.020
		(0.986)	(0.957)	(0.568)	(0.349)	(0.984)		(0.780)	(0.678)	(0.593)			(0.650)
Conflict Intensity		−0.004	−0.068	−0.022	−0.036	−0.004		−0.032	−0.019	−0.006			0.0007
		(0.810)	(0.485)	(0.782)	(0.422)	(0.789)		(0.601)	(0.521)	(0.745)			(0.979)
Conflict Fought		0.582***	0.046	−0.013	0.020	0.0004		−0.012	0.020	−0.010			−0.012
		(0.000)	(0.779)	(0.804)	(0.316)	(0.971)		(0.766)	(0.285)	(0.923)			(0.653)
UNPKF		0.012	0.028	−0.001	0.018*	−0.0001		0.017	0.004	−0.004			0.017
		(0.406)	(0.524)	(0.986)	(0.072)	(0.982)		(0.551)	(0.670)	(0.697)			(0.116)
AR(1)		(0.046)	(0.081)	(0.830)	(0.223)	(0.967)		(0.325)	(0.187)	(0.048)			(0.434)
AR(2)		(0.017)	(0.087)	(0.750)	(0.130)	(0.964)		(0.022)	(0.027)	(0.156)			(0.000)
Sargan OIR		(0.232)	(0.976)	(0.067)	(0.007)	(1.000)		(0.000)	(0.909)	(0.834)			(0.001)
Hansen OIR		(0.363)	(0.999)	(0.996)	(0.968)	(1.000)		(0.746)	(0.896)	(0.989)			(0.773)
DHT for instruments
(a) Instruments in levels H excluding group		(0.593)	(0.387)	(0.570)	(0.504)	(0.871)		(0.117)	(0.549)	(0.462)			(0.373)
Dif (null, H = exogenous)		(0.254)	(1.000)	(1.000)	(0.991)	(1.000)		(0.985)	(0.913)	(1.000)			(0.861)
(b) IV (years, eq(diff)) H excluding group		(0.255)	(0.982)	(0.982)	(0.914)	(1.000)		(0.550)	(0.969)	(0.941)			(0.780)
Dif (null, H = exogenous)		(0.665)	(1.000)	(0.960)	(0.900)	(0.999)		(0.917)	(0.267)	(0.994)			(0.466)
Fisher		21.61***	103.6***	1.68	25.93***	1.27		17.87***	1.96**	1.13			3.66***
Instruments		31	31	31	31	31		31	31	31			31
Countries		48	18	20	44	23		50	87	20			163
Observations		240	90	100	220	115		250	435	100			815

Note. The significance of bold values is twofold. (1) The significance of estimated coefficients and the Fisher statistics. (2) The failure to reject the null hypotheses of (a) no autocorrelation in the AR(1) and AR(2) tests and (b) the validity of the instruments in the Sargan and Hansen OIR tests. AR = Autocorrelation; ECA = Europe and Central Asia; EAP = East Asia and the Pacific; MENA = Middle East and North Africa; SSA = sub-Saharan Africa; LA = Latin America; NA = North America; SA = South Asia; Eng. = English Common Law countries; Frch. = French Civil Law countries; Ger. = German Civil law countries; Scand. = Scandinavian Civil law countries; Social = Socialist countries; UNPKF = United Nations Peace Keeping Force; OIR = Over-identifying Restrictions Test; DHT = Difference in Hansen Test for Exogeneity of Instruments Subsets; Dif = Difference.

, **, ***: significance levels at 10%, 5%, and 1%, respectively.

While the validity of models is relevant for assessing the significance of determinants of deaths from external conflicts, for persistence to be established, other complementary criteria should be met. These complementary criteria are consistent with recent literature from which two conditions are needed for convergence to be established (Fung 2009, 58; Asongu 2013, 192). First, the estimated lagged endogenous variable has to be statistically significant. Second, the statistically significant estimated lagged endogenous variable should be within the interval of zero and one.

It is worth emphasizing why the criterion is based on the lagged endogenous variable instead of the beta coefficient which is traditionally reported in GMM estimation output. When the underlying beta is reported, the criterion for convergence is that it should be less than zero. Alternatively, the estimated lagged dependent variable is directly reported, from which one is subtracted to obtain the underlying beta. To avoid such arithmetic concerns, we directly report the estimated lagged outcome variable and employ the alternative information criterion to assess convergence. This information criterion—which requires that the absolute value corresponding to the estimated lagged outcome variable falls within the suggested interval (i.e., zero and one)—is consistent with recent literature (Asongu and Nwachukwu 2016b, 459).

Before we discuss the results in detail, one final clarification is essential. In the presence of more than one sample within a fundamental characteristic, to establish which sub-sample exhibits more persistence in deaths from internal conflicts, the sub-sample associated with a higher estimated lagged value in deaths from internal conflicts is acknowledged as reflecting more persistence. This is because relevance in magnitude translates the importance of a comparative scope. A higher magnitude implies that past values in deaths from internal conflicts more proportionately affect future values of deaths from internal conflicts.

The following findings can be established from Table 2. First, the deaths from internal conflicts are more persistent in high-income countries compared to low-income countries and, in middle-income countries, the corresponding deaths are more persistent in lower-middle income countries compared to their upper-middle-income counterparts. It follows that our expectation pertaining to cross-country differences based on income levels is only partially valid because it only holds when lower-middle income countries are compared with their upper-middle-income counterparts.

Second, regarding religious domination, persistence of deaths from internal conflicts is apparent in the following order of increasing magnitude: Christian Protestant countries, Christian Catholic countries, and Islam-dominated countries. However, only the findings from Christian Catholic and Islam-oriented countries are used for comparative purposes because of the presence of instrument proliferation in the other specifications (in the post-estimation diagnostic tests, the number of instruments is, respectively, higher than the corresponding number of countries). The comparative tendency in cross sections that pass the instrument proliferation test is not consistent with our earlier discussion because of the persistence in deaths from conflicts that is more apparent in Islam-oriented countries. This unexpected finding can be traceable to the fact that, during the sampled periodicity, most conflicts have been in the MENA due to the Arab Spring (Asongu and Nwachukwu 2016b; El-Haddad 2020). Third, persistence in deaths from internal conflicts is more apparent in coastal countries vis-à-vis their landlocked counterparts. This may be explained by globalization’s tendency to drive conflicts which renders coastal countries more exposed.

In Table 3, results in some sub-samples are not provided because of lack of degrees of freedom. First, from a regional perspective the following regions experience more persistence in deaths from internal conflict in the following increasing order of magnitude: Europe and Central Asia, East Asia and the Pacific, Latin America, SSA, and MENA. Unfortunately, a comparative emphasis is not worthwhile because only the “Europe and Central Asia” sub-sample passes the post-estimation diagnostic test related to instrument proliferation. Second, French civil law countries experience more persistence in deaths compared to their English common law counterparts. This is consistent with our intuitions discussed earlier.

Concerning determinants of persistence in the outcome variable, the following factors are broadly responsible for driving deaths from internal conflicts: homicides, conflict intensity, and conflicts fought. The findings also show that the number of incarcerations deters deaths from internal conflicts. However, on average, the sign of the UNPKF indicator cannot be established with certainty because it reflects both positive and negative effects, contingent on fundamental characteristics.

Further Discussion

Before we discuss these findings further, it is useful to provide more on the concept of convergence underlying the empirical determination of persistence in this study. The analysis is a form of conditional convergence because a conditioning information set is used in the modeling exercise (Asongu 2013). This is contrary to the absolute convergence modeling framework in which only the lagged dependent variable features as an independent variable. Conditional convergence is the type of convergence that shows decreasing dispersions toward a country’s own long-term equilibrium or steady state (Barro 1991; Bruno, de Bonis, and Silvestrini 2012; Narayan, Mishra, and Narayan 2011; Pritchett 1997). Hence, the conditional approach to convergence is contingent on characteristics that determine cross-country differences in the outcome variable. Likewise, the leading roles of the Islam-dominated countries (compared to Christian-oriented countries) and French civil law countries (vis-à-vis English common law countries) are fundamentally traceable to conflict intensity and conflicts fought (which constitute variables or determinants in the conditioning information set). We now offer some explanations to the established tendencies.

The leading role of the Islam-dominated countries in driving deaths from internal conflicts is consistent with events in the post-“2011 Arab Spring” and the periodicity of the study. With respect to the latter, our periodicity employed broadly aligns with the beginning of the Arab Spring. Concerning post-“2011 Arab Spring” events, the conflicts in the MENA region dominated media spotlights since the overthrow of Libyan leader Colonel Muammar Gaddafi. Some features in the MENA region that have contributed to the deaths from internal conflicts include (1) the failed Libyan state in the post-Gaddafi era, in which various rebel factions are fighting to take decisions over the country’s destiny and to determine the law; (2) the scenario in Yemen which is deteriorating due to a proxy war largely fought by Iran and Saudi Arabia who support opposing rebel factions; and (3) the Syrian war which has generated fragile politico-economic environments in neighboring countries (notably, Iraq and Lebanon), mostly because of the birth of the ISIL. Since the MENA countries are Islam-dominated, the narrative above also partly explains why Islam-dominated countries reflect more persistence in deaths from civil conflicts vis-à-vis their Christian-oriented counterparts.

The comparative persistence of French civil law countries vis-à-vis English common law countries can be elicited by the fact that French civil law countries are traditionally less flexible in their ability to adapt to changing environments (Agbor 2015; La Porta et al. 1998, 1999). In light of this insight, the English common law system has advantages in its political and adaptability channels (Beck, Demirgüç-Kunt, and Levine 2013). In the political channel, English common law systems prioritize the rights of individuals instead of the power of the state. In the adaptability mechanism, French civil law countries adapt less to changing circumstances. Therefore, the English common law system is more likely to provide enabling socioeconomic and legal conditions that mitigate the “hysteresis hypothesis” in deaths from conflicts: past observations in deaths from conflicts influence future observations from deaths in conflicts. This perspective can be substantiated by the example of SSA which usefully represents both English common law and French civil law countries. In the post-colonial époque, as of 2014, former French colonies had registered more than half of all documented political coup d’états in Africa—45, versus 22 for English common law countries (Koutonin 2014). These political coup d’états are often surrounded by civil wars and political strife that substantially lead to deaths.

Conclusion and Future Research

This study empirically investigated persistence and determinants of deaths from conflicts around the world from a sample of 163 countries between 2010 and 2015 using the GMM. To permit a discussion on policy implications, the data set was decomposed into fundamentals based on income levels, legal origins, religious domination, landlockedness, and regional proximity. First, the deaths from internal conflicts are more persistent in (1) high-income countries compared to low-income countries, on one hand, and, on the other, in lower-middle-income countries compared to their upper-middle-income counterparts. Second, the persistence of deaths from internal conflict is more apparent in coastal, French civil law, and Islam-oriented countries, compared to landlocked, English common law, Christian-oriented countries, respectively. Third, regarding determinants of persistence, the following factors are significantly responsible for driving deaths from internal conflicts: homicides, conflict intensity, and conflicts fought. We also found that the number of incarcerations have negative effects.

These findings have implications for foreign aid, foreign investment, and tourism location decisions. Given that public spending in developing countries is often not sufficient to prevent and mitigate deaths from conflicts (especially those associated with terrorism), the allocation of foreign aid to developing countries for the purpose of preventing human casualties from conflicts should prioritize those regions (and, by extension, counties) that reflect the highest levels of persistence. In the same vein, tourist arrivals and the location decisions of some foreign direct investments (especially those related to the high traffic of individuals) are most likely to be influenced by evidence of how past deaths from conflicts will influence future deaths from the same cause.

It is also important to note that the established persistence is contingent on the adopted determinants which we have integrated into the models and empirically tested. Future studies can improve the extant literature by focusing on country-specific studies for more targeted policy implications. As we mentioned earlier, to avoid bias and for information, we close this article with the results of this study contained in Tables 4 and 5.

Footnotes

Appendix

Appendix C.

Correlation Matrix (Uniform Sample Size: 977).

S O & P	Homicides	Incar.	IIC	ICF	UNPKF	DFIC	DREC
1.000	−.022	.279	.063	−.092	.0002	.030	−.169	S O & P
	1.000	.162	.296	.184	.305	.170	−.141	Homicides
		1.000	−.082	−.168	−.172	−.142	.066	Incar
			1.000	.528	.301	.542	−.012	IIC
				1.000	.132	.787	.100	ICF
					1.000	.093	−.147	UNPKF
						1.000	.033	DFIC
							1.000	DREC

Note. S O & P = Security Officers & Police; Incar = Incarcerations; IIC = Intensity of Internal Conflict; ICF = Internal Conflict Fought; UNPKF = United Nations Peacekeeping Funding; DFIC = Death from Internal Conflicts; DREC = Death from External Conflicts.

Acknowledgements

The authors would like to thank the editor and anonymous referees at World Affairs whose comments helped to improve the final quality of the text.

ORCID iDs

Simplice A. Asongu

Elda N. Okolo-Obasi

Notes

About the Authors

Simplice A. Asongu holds a PhD from Oxford Brookes University and is currently the Lead Economist and Director of the African Governance and Development Institute (Yaoundé, Cameroon) and Lead Economist and Director of the European Xtramile Centre of African Studies (Liège, Belgium). He is also a senior research fellow at the Africa Growth Institute (Cape Town, South Africa); PhD supervisor at Covenant University in Nigeria, the University of Ghana, and Midlands State University in Zimbabwe; DBA supervisor at the Management College of Southern Africa and research associate at the University of South Africa, University of Buea, and Oxford Brookes University. He can be contacted at asongusimplice@yahoo.com and asongus@afridev.org.

Joseph I. Uduji holds PhD (Marketing), PhD (Public Administration), MSc (Marketing), MSc (Public Relations), MBA (Business Administration), and MPA (Public Administration) from the University of Nigeria, Nsukka, and Enugu Campuses. He lectures in business and development courses in the University of Nigeria and is also a visiting professor to the Catholic University of Cameroon, Bamenda. His current research interests include the impact assessment of multinationals corporate social responsibility in host communities of developing countries. He can be contacted at joseph.uduji@unn.edu.ng and joseph.uduji@gmail.com.

Elda N. Okolo-Obasi holds a BSc (Agricultural Economics) and an MSc (Development Studies) from the University of Nigeria, Nsukka, and Enugu Campuses. He is a doctoral student and research fellow in the Institute for Development Studies, University of Nigeria. His current research interests include economic analysis of multinationals and corporate social responsibility impact in the host communities of developing countries. He can be contacted at eldanduka@yahoo.com and ndukaelda@yahoo.com.

References

Adhikari

Prakash

Hansen

Wendy

Powers

Kathy

. 2012. “The Demand for Reparations: Grievance, Risk, and the Pursuit of Justice in Civil War Settlement.” Journal of Conflict Resolution 56 (2): 183–205.

Agbor

Julius

. 2015. “How Does Colonial Origin Matter for Economic Performance in sub-Saharan Africa?” In Growth and Institutions in African Development, edited by Fosu

Augustin K.

, 309–27. New York: Routledge Studies in Development Economics. https://www.econstor.eu/bitstream/10419/54134/1/657453218.pdf (accessed October 23, 2019).

Akcinaroglu

Seden

Radziszewski

Elizabeth

. 2013. “Private Military Companies, Opportunities, and Termination of Civil Wars in Africa.” Journal of Conflict Resolution 57 (5): 5795–821.

Akdedian

Harout

Hasan

Harith

. 2020. “State Atrophy and the Reconfiguration of Borderlands in Syria and Iraq: Post-2011 Dynamics.” Political Geography 80:102178.

Amadeo

Kimberly

Boyle

Michael

. 2020. “Ukraine Crisis Summary and Explanation.” The Balance, August 21. https://www.thebalance.com/ukraine-crisis-summary-and-explanation-3970462 (accessed August 20, 2020).

Anderson

Mark

. 2015a. “Global Cost of Conflict Reaches $14.3tn, Says Report.” The Guardian, June 17. https://www.theguardian.com/global-development/2015/jun/16/global-cost-conflict-reaches-14tn-says-report#:~:text=The%20world’s%20steady%20shift%20away,Global%20Peace%20Index%20(GPI) (accessed September 30, 2020).

Anderson

Mark

. 2015b. “Nigeria Suffers Highest Number of Civilian Deaths in African War Zones.” The Guardian, January 23. https://www.theguardian.com/global-development/2015/jan/23/boko-haram-nigeria-civilian-death-toll-highest-acled-african-war-zones (accessed August 20, 2020).

Anyanwu

John

Erhijakpor

Andrew

. 2014. “Does Oil Wealth Affect Democracy in Africa?” African Development Review 26 (1): 15–37.

Arellano

Manuel

Bover

Olympia

. 1995. “Another Look at the Instrumental Variable Estimation of Error Components Models.” Journal of Econometrics 68 (1): 29–52.

10.

Arvis

Jean-Francois

Marteau

Jean-Francois

Raballand

Gael

. 2007. “The Cost of Being Landlocked: Logistics Costs and Supply Chain Reliability.” Word Bank Working Paper Series No. 4258. Washington, DC. https://econpapers.repec.org/paper/wbkwbrwps/4258.htm (accessed June 7, 2017).

11.

Asongu

Simplice

. 2012. “Government Quality Determinants of Stock Market Performance in African Countries.” Journal of African Business 13 (2): 183–99.

12.

Asongu

Simplice

. 2013. “African Stock Market Performance Dynamics: A Multidimensional Convergence Assessment.” Journal of African Business 14 (3): 186–201.

13.

Asongu

Simplice

. 2014a. “African Development: Beyond Income Convergence.” South African Journal of Economics 82 (3): 334–53.

14.

Asongu

Simplice

. 2014b. “On the Substitution of Institutions and Finance in Investment.” Economics Bulletin 34 (3): 1557–74.

15.

Asongu

Simplice

. 2019. “Persistence in Incarcerations: Global Comparative Evidence.” Journal of Criminological Research, Policy and Practice 4 (2): 136–47.

16.

Asongu

Simplice.

2020. “The Persistence of Global Terrorism.” Territory, Politics, Governance. Published electronically November. doi:10.1080/21622671.2019.1675532.

17.

Asongu

Simplice

Acha-Anyi

Paul

. 2019. “The Murder Epidemic: A Global Comparative Study.” International Criminal Justice Review 29 (2): 105–20.

18.

Asongu

Simplice

Acha-Anyi

Paul

. 2020. “Global Tourism and Waves of Terror: Perspectives from Military Expenditure.” Journal of Policy Research in Tourism, Leisure and Events 12 (2): 239–61.

19.

Asongu

Simplice

De Moor

Lieven

. 2017. “Financial Globalisation Dynamic Thresholds for Financial Development: Evidence from Africa.” European Journal of Development Research 29 (1): 192–212.

20.

Asongu

Simplice

Kodila-Tedika

Oasis

. 2016. “Fighting African Conflicts and Crimes: Which Governance Tools Matter?” International Journal of Social Economics 43 (5): 466–85.

21.

Asongu

Simplice

Kodila-Tedika

Oasis

. 2017. “Trade, Aid and Terror.” International Journal of Development Issues 16 (1): 2–24.

22.

Asongu

Simplice

le Roux

Sara

. 2017. “Enhancing ICT for Inclusive Human Development in Sub-Saharan Africa.” Technological Forecasting and Social Change 118:44–54.

23.

Asongu

Simplice

Nnanna

Joseph

. 2019a. “Dynamic Determinants of Access to Weapons: Global Evidence.” Foreign Trade Review 54 (4): 334–54.

24.

Asongu

Simplice

Nnanna

Joseph

. 2019b. “Foreign Aid, Instability, and Governance in Africa.” Politics & Policy 47 (4): 807–48.

25.

Asongu

Simplice

Nnanna

Joseph

Biekpe

Nicholas

Acha-Anyi

Paul

. 2019. “Contemporary Drivers of Global Tourism: Evidence from Terrorism and Peace Factors.” Journal of Travel & Tourism Marketing 36 (3): 345–57.

26.

Asongu

Simplice

Nwachukwu

Jacinta

. 2016a. “Foreign Aid and Governance in Africa.” International Review of Applied Economics 30 (1): 69–88.

27.

Asongu

Simplice

Nwachukwu

Jacinta

. 2016b. “Revolution Empirics: Predicting the Arab Spring.” Empirical Economics 51 (2): 439–82.

28.

Asongu

Simplice

Nwachukwu

Jacinta

. 2016c. “The Mobile Phone in the Diffusion of Knowledge for Institutional Quality in Sub-Saharan Africa.” World Development 86:133–47.

29.

Asongu

Simplice

Nwachukwu

Jacinta

. 2016d. “The Role of Governance in Mobile Phones for Inclusive Human Development in Sub-Saharan Africa.” Technovation 55–56:1–13.

30.

Asongu

Simplice

Nwachukwu

Jacinta

. 2016e. “The Role of Lifelong Learning in Political Stability and Non-violence: Evidence from Africa.” Journal of Economic Studies 43 (1): 141–64.

31.

Asongu

Simplice

Nwachukwu

Jacinta

. 2017a. “Mobile Phones in the Diffusion of Knowledge and Persistence in Inclusive Human Development in Sub-Saharan Africa.” Information Development 33 (3): 289–302.

32.

Asongu

Simplice

Nwachukwu

Jacinta

. 2017b. “Quality of Growth Empirics: Comparative Gaps, Benchmarking and Policy Syndromes.” Journal of Policy Modeling 39 (5): 861–82.

33.

Asongu

Simplice

Nwachukwu

Jacinta

. 2017c. “The Comparative Inclusive Human Development of Globalisation in Africa.” Social Indicators Research 134 (3): 1027–50.

34.

Asongu

Simplice

Nwachukwu

Jacinta

. 2017d. “The Impact of Terrorism on Governance in African Countries.” World Development 99:253–70.

35.

Asongu

Simplice

Nwachukwu

Jacinta

Pyke

Chris

. 2019. “The Right to Life: Global Evidence on the Role of Security Officers and the Police in Modulating the Effect of Insecurity on Homicide.” Social Indicators Research 143 (2): 727–40.

36.

Asongu

Simplice

Tchamyou

Vanessa

Asongu

Ndemaze

Tchamyou

Nina

. 2018. “The Comparative African Economics of Governance in Fighting Terrorism.” African Security 11 (4): 296–338.

37.

Asongu

Simplice

Tchamyou

Vanessa

Minkoua

Jules

Asongu

Ndemaze

Tchamyou

Nina

. 2018. “Fighting Terrorism in Africa: Benchmarking Policy Harmonization.” Physica A: Statistical Mechanics and Its Applications 492:1931–57.

38.

Asongu

Simplice

Uduji

Joseph

Okolo-Obasi

Elda

. 2019. “Tourism and Insecurity in the World.” International Review of Economics 66 (4): 453–72.

39.

Asongu

Simplice

Uduji

Joseph

Okolo-Obasi

Elda

. 2020a. “The Persistence of Weapons: Global Evidence.” Politics & Policy 48 (1): 191–224.

40.

Asongu

Simplice

Uduji

Joseph

Okolo-Obasi

Elda

. 2020b. “Political Instability and Political Terror: Global Evidence on Persistence.” Journal of Public Affairs 20 (3): e2119.

41.

Baltagi

Badi

. 2008. “Forecasting with Panel Data.” Journal of Forecasting 27 (2): 153–73.

42.

Barro

Robert

. 1991. “Economic Growth in a Cross Section of Countries.” Quarterly Journal of Economics 106 (2): 407–43.

43.

Barro

Robert

Sala-I-Martin

Xavier

. 1992. “Convergence.” Journal of Political Economy 100 (2): 223–51.

44.

Barro

Robert

Sala-I-Martin

Xavier

. 1995. Economic Growth. Cambridge: The MIT Press.

45.

Baumol

William

. 1986. “Productivity, Growth, Convergence and Welfare: What the Long Run Data Show.” American Economic Review 78 (5): 1072–85.

46.

BBC News . 2020. “Myanmar Rohingya: What You Need to Know about the Crisis.” January 23. https://www.bbc.com/news/world-asia-41566561 (accessed July 13, 2020).

47.

Beck

Thorsten

Demirgüç-Kunt

Asli

Levine

Ross

. 2003. “Law and Finance: Why Does Legal Origin Matter?” Journal of Comparative Economics 31 (4): 653–75.

48.

Beegle

Kathleen

Christiaensen

Luc

Dabalen

Andrew

Isis

Gaddis

. 2016. Poverty in a Rising Africa. Washington, DC: The World Bank.

49.

Ben-Meir

Alon

. 2020. “Middle East: Ever More Unstable.” The Globalist, January 5. https://www.theglobalist.com/middle-east-iran-syria-yemen-war-iraq-israeli-palestinian-conflict/ (accessed July 13, 2020).

50.

Blackburn

Christine

Lenze

Paul

Casey

Rachel

. 2020. “Conflict and Cholera: Yemen’s Man-Made Public Health Crisis and the Global Implications of Weaponizing Health.” Health Security 8 (2): 125–31.

51.

Boateng

Agyenim

Asongu

Simplice

Akamavi

Raphael

Tchamyou

Vanessa

. 2018. “Information Asymmetry and Market Power in the African Banking Industry.” Journal of Multinational Financial Management 44:69–83.

52.

Bodine

Richard

Crawford

Donna

. 1998. The Handbook of Conflict Resolution Education: A Guide to Building Quality Programs in Schools. San Francisco: Jossey-Bass. https://www.wiley.com/en-us/The+Handbook+of+Conflict+Resolution+Education%3A+A+Guide+to+Building+Quality+Programs+in+Schools-p-9780787910969 (accessed October 23, 2019).

53.

Brancati

Dawn

Snyder

Jack

. 2013. “Time to Kill the Impact of Election Timing on Post Conflict Stability.” Journal of Conflict Resolution 57 (5): 822–53.

54.

Bruno

Giuseppe

de Bonis

Riccardo

Silvestrini

Andrea

. 2012. “Do Financial Systems Converge? New Evidence from Financial Assets in OECD Countries.” Journal of Comparative Economics 40 (1): 141–55.

55.

Central Intelligence Agency. 2011. “The World Factbook Download 2011.” https://www.cia.gov/library/publications/download/download-2011/ (accessed October 23, 2019).

56.

Choi

Seung-Whan

. 2010. “Fighting Terrorism through the Rule of Law?” The Journal of Conflict Resolution 54 (6): 940–66.

57.

Costa

Albert

Hermandez

Mireia

Sebastian-Gallés

Nuria

. 2008. “Bilingualism Aids Conflict Resolution: Evidence from the ANT Task.” Cognition 106 (1): 59–86.

58.

D’Amico

Daniel

. 2010. “The Prison in Economics: Private and Public Incarceration in Ancient Greece.” Public Choice 145 (3–4): 461–82.

59.

Darwich

May

. 2020. “Escalation in Failed Military Interventions: Saudi and Emirati Quagmires in Yemen.” Global Policy 11 (1): 103–12.

60.

David

Anda

Marouani

Mohamed

Nahas

Charbel

Nilsson

Bjorn

. 2020. “The Economics of the Syrian Refugee Crisis in Neighbouring Countries: The Case of Lebanon.” Economics of Transition and Institutional Change 28 (1): 89–109.

61.

Doyle

Joanne

. 2017. “Persistence in the Long-run Expected Rate of Return for Corporate Pension Plans.” Quarterly Review of Economics and Finance 63:271–77.

62.

Drakos

Konstantinos

Gofas

Andreas

. 2006. “In Search of the Average Transnational Terrorist Attack Venue.” Defence and Peace Economics 17 (2): 73–93.

63.

Druce

Stephen

. 2020. “Myanmar’s Unwanted Ethnic Minority: A History and Analysis of the Rohingya Crisis.” In Managing Conflicts in a Globalizing ASEAN, edited by Oishi

Singapore: Springer. doi:10.1007/978-981-32-9570-4_2.

64.

Efobi

Uchenna

. 2015. “Politicians’ Attributes and Institutional Quality in Africa: A Focus on Corruption.” Journal of Economic Issues 49 (3): 787–813.

65.

Efobi

Uchenna

Asongu

Simplice

Beecroft

Ibukun

. 2015. “Foreign Direct Investment, Aid and Terrorism: Empirical Insight Conditioned on Corruption Control.” African Governance and Development Institute Working Paper No. 15/007. https://ideas.repec.org/p/abh/wpaper/15-007.html (accessed October 23, 2019).

66.

El-Haddad

Amirah

. 2020. “Redefining the Social Contract in the Wake of the Arab Spring: The Experiences of Egypt, Morocco and Tunisia.” World Development 127:104774.

67.

Eubank

William

Weinberg

Leonard

. 1994. “Does Democracy Encourage Terrorism?” Terrorism and Political Violence 6 (4): 417–43.

68.

Franco

Annie

Malhotra

Neil

Simonovits

Gabor

. 1991. “Publication Bias in the Social Sciences: Unlocking the File Drawer.” Department of Political Science, Stanford University, Stanford. https://science.sciencemag.org/content/345/6203/1502 (accessed December 28, 2014).

69.

Fung

Michael

. 2009. “Financial Development and Economic Growth: Convergence or Divergence?” Journal of International Money and Finance 28 (1): 56–67.

70.

Gaibulloev

Khusrav

Sandler

Todd

. 2009. “The Impact of Terrorism and Conflicts on Growth in Asia.” Economics and Politics 21 (3): 359–83.

71.

Goddard

John

Liu

Hong

Molyneux

Philip

Wilson

John

. 2011. “The Persistence of Bank Profit.” Journal of Banking & Finance 35 (11): 2881–90.

72.

Gubler

Joshua

Selway

Joe

. 2012. “Horizontal Inequality, Crosscutting Cleavages, and Civil War.” Journal of Conflict Resolution 56 (2): 206–32.

73.

Islam

. 2020. “Communal Peace in Bangladesh after Rohingya Crisis: An Assessment of Different Approaches to Community Development.” Journal of Prevention & Intervention in Community 48 (3): 256–71.

74.

Hyeran

Namgung

Hyun

. 2012. “Dispute Settlement Mechanisms in Preferential Trade Agreements Democracy, Boilerplates, and the Multilateral Trade Regime.” Journal of Conflict Resolution 56 (6): 1041–68.

75.

Justin

Peter

de Vries

Lotje

. 2019. “Governing Unclear Lines: Local Boundaries as a (Re)source of Conflict in South Sudan.” Journal of Borderlands Studies 34 (1): 31–46.

76.

Keke

Reginald

. 2020. “Southern Cameroons/Ambazonia Conflict: A Political Economy.” Theory & Event 23 (2): 329–51.

77.

Kossele

Pacific

. 2020. “Fragility of State in Central African Republic: An Econometric Approach to Efficiency Understanding.” Global Business Review 21 (3): 681–97.

78.

Kossele

Pacific

Shan

. 2018. “Economic Security and the Political Governance Crisis in Central African Republic.” African Development Review 30 (4): 462–77.

79.

Koutonin

Mawuna

. 2014. “14 African Countries Forced by France to Pay Colonial Tax for the Benefits of Slavery and Colonization.” Silicon Africa. https://blogs.mediapart.fr/jecmaus/blog/300114/franceafrique-14-african-countries-forced-france-pay-colonial-tax-benefits-slavery-and-colonization#:~:text=%232.&text=France%20has%20been%20holding%20the,Brazzaville%2C%20Equatorial%20Guinea%20and%20Gabon

80.

Krause

Jana

. 2019. “Stabilization and Local Conflicts: Communal and Civil War in South Sudan.” Ethnopolitics 18 (5): 478–93.

81.

Lacina

Bethany

. 2006. “Explaining the Severity of Civil Wars.” Journal of Conflict Resolution 50 (2): 276–89.

82.

Lacina

Bethany

Gleditsch

Nils

. 2005. “Monitoring Trends in Global Combat: A New Dataset of Battle Deaths.” European Journal of Population 21 (2–3): 145–66.

83.

Lacina

Bethany

Gleditsch

Nils

Russett

Bruce

. 2006. “The Declining Risk of Death in Battle.” International Studies Quarterly 50 (3): 673–80.

84.

Lai

Brian

. 2007. “‘Draining the Swamp’: An Empirical Examination of the Production of International Terrorism, 1968–1998.” Conflict Management and Peace Science 24 (4): 297–310.

85.

Laipson

Ellen

. 2019. “Iran and the Nuclear Agreement: Stalemate or on the Road to War?” SAIS Review of International Affairs 39 (2): 119–27.

86.

La Porta

Rafael

Lopez-De-Silanes

Florencio

Shleifer

Andrei

. 2008. “The Economic Consequences of Legal Origin.” Journal of Economic Literature 46 (2): 285–332.

87.

La Porta

Rafael

Lopez-De-Silanes

Florencio

Shleifer

Andrei

Vishny

Robert

. 1998. “Law and Finance.” Journal of Political Economy 106 (6): 1113–55.

88.

La Porta

Rafael

Lopez-De-Silanes

Florencio

Shleifer

Andrei

Vishny

Robert

. 1999. “The Quality of Government.” Journal of Law, Economics and Organization 15 (1): 222–79.

89.

Lenski

Tammy

. 2014. The Conflict Pivot: Turning Conflict into Peace of Mind. Peterborough: MyriaccordMedia.

90.

Quan

. 2005. “Does Democracy Promote Transnational Terrorist Incidents?” Journal of Conflict Resolution 49 (2): 278–97.

91.

Quan

Wen

Ming

. 2005. “The Immediate and Lingering Effects of Armed Conflict on Adult Mortality: A Time-series Cross-national Analysis.” Journal of Peace Research 42 (4): 472–91.

92.

Love

Inessa

Zicchino

Lea

. 2006. “Financial Development and Dynamic Investment Behavior: Evidence from Panel VAR.” The Quarterly Review of Economics and Finance 46 (2): 190–210.

93.

Luhnow

David

. 2018. “Latin America Is the Murder Capital of the World.” The Wall Street Journal, September 20. https://www.wsj.com/articles/400-murders-a-day-the-crisis-of-latin-america-1537455390 (accessed September 15, 2020).

94.

Malhotra

Deepak

. 2016. Negotiating the Impossible: How to Break Deadlocks and Resolve Ugly Conflicts (Without Money or Muscle). Oakland: Berrett-Koehler.

95.

Mankiw

Gregory

Romer

David

Weil

David

. 1992. “A Contribution to the Empirics of Economic Growth.” Quarterly Journal of Economics 107 (2): 407–37.

96.

Mayer-Foulkes

David

. 2010. “Divergences and Convergences in Human Development.” UNDP Human Development Research Paper No. 20/2010. http://hdr.undp.org/en/content/divergences-and-convergences-human-development (accessed October 23, 2019).

97.

Mbih

Richard.

2020. “The Politics of Farmer–Herder Conflicts and Alternative Conflict Management in Northwest Cameroon.” African Goegraphical Review. Published electronically January. doi:10.1080/19376812.2020.1720755.

98.

Mlachila

Montfort

Tapsoba

Rene

Tapsoba

Sampawende

. 2017. “A Quality of Growth Index for Developing Countries: A Proposal.” Social Indicators Research 134 (2): 675–710.

99.

Muggah

Robert

de Carvalho

Ilona

. 2017. “There’s a Cure for Latin America’s Murder Epidemic—and It Doesn’t Involve More Police or Prisons.” World Economic Forum. https://www.weforum.org/agenda/2017/04/there-s-a-cure-for-latin-america-s-murder-epidemic-and-it-doesn-t-involve-more-police-or-prisons/ (accessed September 30, 2020).

100.

Murray

Christopher

King

Garry

Lopez

Alan

Tomijima

Niels

Krug

Etienne

. 2002. “Armed Conflict as a Public Health Problem.” British Medical Journal 324:346–49.

101.

Narayan

Paresh

Mishra

Sangarika

Narayan

Seema

. 2011. “Do Market Capitalization and Stocks Traded Converge? New Global Evidence.” Journal of Banking and Finance 35 (10): 2771–81.

102.

Nganji

Julius

Cockburn

Lynn

. 2020. “Use of Twitter in the Cameroon Anglophone Crisis.” Behavior & Information Technology 39 (3): 267–87.

103.

Obermeyer

Ziad

Murray

Christopher

Gakidou

Emmanuela

. 2008. “Fifty Years of Violent War Deaths from Vietnam to Bosnia: Analysis of Data from the World Health Survey Programme.” British Medical Journal 336:1482.

104.

Pelizzo

Ricardo

Araral

Eduardo

Pak

Anton

Xun

. 2016. “Determinants of Bribery: Theory and Evidence from Sub-Saharan Africa.” African Development Review 28 (2): 229–40.

105.

Pelizzo

Ricardo

Nwokora

Zim

. 2016. “Bridging the Divide: Measuring Party System Change and Classifying Party Systems.” Politics & Policy 44 (6): 1017–52.

106.

Pelizzo

Ricardo

Nwokora

Zim

. 2018. “Party System Change and the Quality of Democracy in East Africa.” Politics & Policy 46 (3): 505–28.

107.

Piazza

James

. 2007. “Draining the Swamp: Democracy Promotion, State Failure, and Terrorism in 19 Middle Eastern States.” Studies in Conflict and Terrorism 30 (6): 521–39.

108.

Piazza

James

. 2008. “Incubators of Terror: Do Failed and Failing States Promote Transnational Terrorism?” International Studies Quarterly 52 (3): 468–88. https://www.jstor.org/stable/29734247 (accessed October 23, 2019).

109.

Pieper

Moritz

. 2019. “An Iran Nuclear Deal without the United States? Chinese, European, and Russian Interests and Options after the US Withdrawal from the Joint Comprehensive Plan of Action.” In Contemporary Issues in International Political Economy, edited by Yu

F. L.

Kwan

Singapore: Palgrave Macmillan. doi:10.1007/978-981-13-6462-4_2.

110.

Pritchett

Lant

. 1997. “Divergence, Big Time.” Journal of Economic Perspectives 11 (3): 3–17.

111.

Roodman

David

. 2009a. “A Note on the Theme of too Many Instruments.” Oxford Bulletin of Economics and Statistics 71 (1): 135–58.

112.

Roodman

David

. 2009b. “How to do Xtabond2: An Introduction to Difference and System GMM in Stata.” Stata Journal 9 (1): 86–136.

113.

Rosenberg

Michael

. 2005. “The File-drawer Problem Revisited: A General Weighted Method for Calculating Fail-safe Numbers in Meta-analysis.” Evolution 59 (2): 464–68.

114.

Rubenstein

Richard

. 2017. Resolving Structural Conflicts: How Violent Systems Can Be Transformed. London: Routledge.

115.

Schmid

Alex

. 1992. “Terrorism and Democracy.” Terrorism and Political Violence 4 (4): 14–25.

116.

Searcey

Dionne

Santora

Marc

. 2015. “Boko Haram Ranked Ahead of ISIS as Deadliest Terror Group.” The New York Times, November 19. https://www.nytimes.com/2015/11/19/world/africa/boko-haram-ranked-ahead-of-isis-for-deadliest-terror-group.html?mcubz=0 (accessed October 23, 2019).

117.

Singh

Bilveer

. 2020. “Implications of Soleimani’s Killing for South and Southeast Asia.” Counter Terrorist Trends and Analyses 12 (2): 12–16.

118.

Stephan

Andreas

Tsapin

Andriy

. 2008. “Persistence and Determinants of Firm Profit in Emerging Markets.” Applied Economics Quarterly 54 (4): 231–53.

119.

Tchamyou

Vanessa

. 2019. “The Role of Information Sharing in Modulating the Effect of Financial Access on Inequality.” Journal of African Business 20 (3): 317–38.

120.

Tchamyou

Vanessa

. 2020. “Education, Lifelong learning, Inequality and Financial Access: Evidence from African Countries.” Contemporary Social Science 15 (1): 7–25.

121.

Tchamyou

Vanessa

Erreygers

Guido

Cassimon

Dany

. 2019. “Inequality, ICT, and Financial Access in Africa.” Technological Forecasting and Social Change 139:169–84.

122.

The Economist . 2017. “The World’s Most Dangerous Cities.” March 31. https://www.economist.com/graphic-detail/2017/03/31/the-worlds-most-dangerous-cities (accessed September 15, 2020).

123.

Walter

Barbara

. 2015. “Why Bad Governance Leads to Repeat Civil War.” Journal of Conflict Resolution 59 (7): 1242–72.

124.

Walther

Olivier

Perdersen

Patrick

. 2020. “Rebel Fragmentation in Syria’s Civil War.” Small.” Wars and Insurgencies 31 (3): 445–74.

125.

Zarei

Mohanna

. 2020. “The Water-energy-food Nexus: A Holistic Approach for Resource Security in Iran, Iraq, and Turkey.” Water-Energy Nexus 3:81–94.