You Might Have Missed: Academic Journals IV

This is the fourth blog post in this series. The previous three were published in February, July, and October 2015, and highlight earlier academic findings. 

Cullen S. Hendriz, “A House Divided: Threat Perception, Military Factionalism, and Repression in Africa,” Journal of Conflict Resolution, online, January 7, 2016.

We find that military factionalism, as measured by past intraregime violence, is associated with a significant reduction in the likelihood of repression….We do not find that economic claims and political claims have a statistically significant interactive effect with past military factionalism.

We find also that ethnic/religious claims of protesters have a negative and statistically significant interactive effect with regime factionalization. When contentious events involve ethnic or religious claims, countries with a history of past intragovernment violence are especially less likely to conduct repression. (p. 13)

At zero instances of past intraregime violence, ethnic/religious dissent is 25 percent more likely to be repressed than dissent which is not ethnically/religiously motivated….However, as past intraregime violence increases, ethnically/religiously motivated events are less likely to be repressed. In the extreme, a country with ten previous instances of intraregime violence is 47 percent less likely to repress an ethnically/religiously motivated event than a nonethnically/ religiously motivated event….When past factionalism is not present, governments behave more repressively to ethnically and religiously motivated challenges. However, when there is a history of past military factionalism, these challenges are much less likely to be repressed.

Further investigation of this dynamic reveals an interesting interaction with a third component of threat perception: whether the central government is directly targeted. Even if the military is characterized by a history of factionalism and ethnic/religious claims are being made, these claims may not be perceived as especially threatening by the government if they do not target the government. (p. 16)

We demonstrate that while ethnically and religiously motivated challenges are more frequently met with repression in cohesive regimes, past intraregime violence makes leaders much less likely to repress these challenges. Given that many African militaries are characterized by ethnic cleavages, popular unrest along ethnic lines is especially likely to split the regime. For this reason, leaders will be particularly reluctant to engage in repression. (p. 22)

Patrick B. Johnston and Anoop K. Sarbahi, “The Impact of US Drone Strikes on Terrorism in Pakistan,” International Studies Quarterly, online, January 5, 2016.

We find no statistically significant evidence of a positive relationship between drone strikes and terrorism. The results instead support the alternative thesis as elaborated in Hypothesis 2: drone strikes are associated with decreases in militant violence…

These results suggest that the unpopularity of drone strikes is associated with relatively high levels of civilian support for militancy and increased anti-Americanism. (pp. 8-9)

Given that drone strikes are associated with reductions in militant attacks in the areas where they occur, we also expect drone strikes to be negatively associated with the lethality, or “quality,” of militant attacks in these same areas. This is indeed the case….The lethality of militant attacks declined by an average of nearly 25 percentage points in a given week in which a drone strike occurred…

The evidence is consistent with observable implications of a “disruption” mechanism, suggesting that the threat to militants posed by drone strikes inhibits insurgent and terrorist groups from conducting operational activities at the same rate at which they are able to perpetrate such activities in the absence of drone strikes…

The death of a HVI [high-value individual] is statistically associated with a decrease in terrorist attacks….However, the results are not statistically significant. (p. 10)

Vincent Bauer, Keven Ruby, and Robert Pape, “Solving the Problem of Unattributed Political Violence,” Journal of Conflict Resolution, 2015, pp. 1-28.

Based on attacks in FATA, drone strikes do not have an effect on the level of Pakistani Taliban violence. The model was not confident that the number of attacks changed in either direction for the first four weeks following a drone strike but was confident that the change in attacks was close to zero in the next four-week period.

In contrast, based on the full range of Pakistani Taliban operations attributed by MI, drone strikes consistently suppress militant violence by about 5 percent over the entire eight-week period, or about one fewer attack per week. Using a larger sample of attacks significantly improved the model’s confidence in comparison to using claimed attacks alone, as illustrated by the narrower confidence bounds, and very little of the remaining uncertainty was due to variation between the imputed data sets. In terms of policy, the findings produced by the MI approach suggest that drone strikes are only marginally effective at reducing militant violence in the short term, and that the effect dissipates over time. (p. 23)

Todd C. Robinson, “What Do We Mean by Nuclear Proliferation?” The Nonproliferation Review, 22:1, 2015, pp. 53-70.

An analysis of these two conceptualizations—attempted acquisition/development and increased capability—illustrates one of the principal issues plaguing much of the nuclear proliferation research: many of its conceptualizations are logically inconsistent. An increase in the size and/or sophistication of an existing nuclear arsenal, which implies that a state actually possesses nuclear weapons, and the development of production capabilities are not simply constituent parts of the same general phenomenon, as competitiveness and openness are for democracy; they are logically inconsistent, in that a state cannot both be developing nuclear weapons production capabilities and possess nuclear weapons. What proliferation means, therefore, remains unclear. (p. 57)

There are two ways in which this conceptualization can be interpreted. The first is that proliferation is the process by which acquisition occurs. Thus, as a delimiting mechanism, anything that occurs either before or after this process is not proliferation. The second is that proliferation is a result of acquisition, i.e. it occurs only when acquisition concludes. Here, what distinguishes proliferation from not proliferation is the actual physical obtainment of nuclear weapons, rather than the process by which this occurs. Obviously, proliferation cannot be both acquisition and the process by which acquisition occurs; yet it is routinely used to describe both. From a policy-making perspective, this is particularly relevant when attempting to determine at what point intervention with the purposes of preventing the occurrence of proliferation should occur. (pp. 58-59)

If we accept the argument that proliferation is not an objective thing, then it is necessarily subjective: proliferation, one could thus argue, does not objectively occur, it is perceived. It is, therefore, subject to bias, informed by any number of factors, including the observer’s relationship with the alleged proliferator and the observer’s own history with nuclear proliferation. (p. 66)

Cigdem V. Sirin and Michael T. Koch, “Dictators and Death: Casualty Sensitivity of Autocracies in Militarized Interstate Disputes,” International Studies Quarterly, 59:4, December 2015, pp. 802-814.

Why are some authoritarian regimes so quick to surrender amid lower numbers of casualties while others prove willing to incur significant casualty counts to continue their war efforts? In this study, we explore the propensity of different authoritarian regime types to sustain casualties in interstate conflicts….First, we employ an inverse divide-the-dollar game (with respect to the distribution of public “bads”) in order to develop a theoretical model that examines casualty sensitivity among autocracies. Second, we analyze the number of casualties different autocratic regime types sustain in militarized interstate disputes (MIDs) [the threat, display or use of military force short of war by one state explicitly directed toward another state.] during the period from 1946 to 2001. (p. 803)

Single-party regimes sustain about 14,881 less casualties and military regimes sustain about 32,282 less casualties as compared to personalist regimes…

The results show that autocracies with larger populations actually have lower numbers of casualties. By comparison, autocracies that have larger militaries are likely to sustain higher casualties….Furthermore, the initiation and duration variables both display positive and moderately significant effects on the number of casualties, suggesting that casualties tend to be higher if the autocratic regime is the initiator of the dispute and when it is involved in protracted militarized conflicts…

Compared to personalist regimes, single-party regimes are significantly less likely to engage in militarized disputes with casualties. Military regimes do not significantly differ from personalist ones in their propensity to get involved in disputes where at least some casualties are incurred. However, both single-party and military regimes are significantly less likely to engage in high-casualty disputes than personalist regimes. (p. 808)

Sebastian Jackle and Marcel Baumann, “‘New Terrorism’ = Higher Brutality? An Empirical Test of the ‘Brutalization Thesis’,” Terrorism and Political Violence, November 5, 2015.

The average number of fatalities due to terrorist attacks rose to a volatile but high level in the 1990s, which it kept until it fell again in 2008 to a level below the one from the 1980s. This indicator thus shows signs of a brutalization. Yet, the starting point for this development was clearly not 9/11 but about a decade before 2001…

In the beginning of the 1990s, for the first time, more persons died in attacks against soft targets than in attacks against hard targets. Yet, this finding can be traced back to three periods in which there was an extremely high death toll amongst the civilian population: 1997, 2001, and 2006 to 2008….All three peaks can be explained either by the extreme case of the 9/11 attacks or by specific, local (civil) war configurations, which makes the distinction between fatalities due to warfare and those due to terrorism extremely difficult. Thus, these periods are not very well-suited to illustrate a general trend towards a brutalization in terrorism. Summing up, if we may talk of a brutalization based on the distinction between soft and hard targets, it has certainly not started with the 9/11 attacks but rather in the early 1990s.

The remaining two indicators are the only ones backing the idea that 9/11 could be seen as a starting point for a more brutal terrorism. The first is the number of suicide bombings, which grew strongly in the years after the U.S. invasion of Afghanistan. Brutalization according to this indicator cannot only be seen from the perspective of the victims, but also from the perspective of the individual terrorists themselves, who are exploited in the most fundamental way by the terrorist organizations behind suicide bombings. The second one is the number of beheadings, which has grown extensively after 2003. Both indicators are closely connected to certain countries (Afghanistan, Iraq, Pakistan) and Islamist terrorist groups. They both emphasize changes in the terrorists’ strategies, which focus more and more towards publicity and media attention. These two indicators thus can indeed be regarded as a qualitative difference compared to earlier decades; however, it is questionable if they justify speaking of a new era of “mega-terrorism” or “catastrophic terrorism,” which would have led to an entirely new quality of global terrorism. (pp. 23-24)

Johann Park and Valentina Bali, “International Terrorism and the Political Survival of Leaders,” The Journal of Conflict Resolution, November 3, 2015.

Terrorism has a statistically significant positive impact on Exit [of a political leader] in line with Hypothesis 1, suggesting that as the number of terror attacks increases, the likelihood that an incumbent leader loses power increases…

It is not fatal incidents but nonfatal ones that significantly reduce the spell of leaders in office…As we go from no incident…to one incident…the probability of leader exit increases by about 8.4 percent. Going from one incident…to five incidents…increases the probability of leader removal by about 13.5 percent. An increase from the fifth percentile value, no incidents…to the ninety-fifth percentile value, eleven attacks…produces a 33.2 percent increase in leader removal…

Interstate war hardly affects the political survival. Not a single war dummy is significant in any model. This may be because leaders tend to go to war only when they know it does not destabilize their political power. Civil war appears to increase the likelihood that political leaders lose their office. Yet, its statistical significance is not robust across specifications....Indeed, of all the threat-related correlates, it is only Terrorism that consistently affects leadership change ….Leaders in wealthy countries seem politically safer than those in poor countries. (p. 13)

Ronald R. Krebs, “How Dominant Narratives Rise and Fall: Military Conflict, Politics, and the Cold War Consensus,” International Organization, 69:4, Fall 2015, pp. 809-845.

Based on a large-scale content analysis of newspaper editorials on foreign affairs, this article shows that the Cold War narrative was narrower than conventional accounts suggest, that it did not coalesce until well into the 1950s, and that it began to erode even before the Vietnam War’s Americanization in 1965. (p. 809)

First, there was a dominant Cold War narrative, or consensus, in the 1950s, but it was narrow. The two newspapers’ [the New York Times and Chicago Tribune] editorials largely moved in lockstep regarding the centrality of superpower competition for global politics; on this issue, the gap between them was small and stable.…They agreed that the communist world was monolithic, that leading communist powers were expansionist and aggressive, that the threat was real, that relations between the two blocs were generally conflictual, and that the domino theory held….Throughout the 1950s and early 1960s, the two papers’ editorials diverged over the reliability of allies, over whether the United States had special responsibilities as leader of the free world, over the value of foreign assistance, and even over whether the United States had in the past acted, or would in the future act, as a force for good in world affairs.

Second, the usual periodization of even the narrow Cold War consensus is problematic: the consensus did not consolidate until well into the Korean War. (pp. 816-817)

Third, the two papers’ articulated views did eventually diverge…from the early- to mid-1960s through the end of the Cold War, statistically significant gaps arose between the two newspapers on every question that had previously constituted the dominant narrative. The Tribune hewed to the traditional Cold War narrative, hardly diverging from its 1950s stance, whereas the Times deviated substantially. Those differences were substantively significant as well, averaging above 17 percent. But they emerged before leading figures in Congress, and before the American people, opposed the Vietnam War, and even before the Americanization of the war in 1965….As the Vietnam War dragged on, there was even less agreement on representations of the

Erin M. Kearns, Brendan Conlon, and Joseph K. Young, “Lying About Terrorism,” Studies in Conflict & Terrorism, 37:5, 2014, pp. 422-439.

During the past few decades, the percentage of claimed attacks has dropped from 61 percent in the 1970s to just 14.5 percent from the late 1990s through 2004. Credit is no longer taken for many of the most deadly attacks. (p. 423)

A group can lie about a terrorist attack that they did not perpetrate in two ways as well: by taking credit for another group’s attack and by blaming it on another group without knowing what group was truly responsible. The levels of internal and external control exerted by a group may help determine the type of lie that the group chooses to tell. Each of these lies about terrorism can be strategically employed to help a group achieve its desired goal(s)….Traditionally, groups have perpetrated terrorism because they want publicity, not a high body count; however, recent attacks are more lethal and less likely to be claimed than attacks in the past. (p. 424)

First, a group may claim responsibility for an attack that they did not perpetrate. Second, a group may commit false flag terrorism by carrying out an attack and then blaming it on a rival organization. Third, a group may lie by blaming an attack that it did not commit on a rival group, which is termed the hot-potato problem. Fourth, a group may engage in a lie of omission where they perpetrate an attack but neither claim responsibility for it nor blame it on another group. (p. 425)

The group that falsely claims an attack may be trying to convince its target of persuasion that it is a credible threat. Attrition as a strategic logic of terrorism is used when a group’s power to make good on its threats is questioned. A group could take credit for another group’s attack to convince the enemy of its power. A group that falsely claims another’s attack could lack the power it is attempting to display, or it could have the capacity to carry out the attack itself but opportunistically claim credit to prevent another group from demonstrating its power….However, when there are multiple competing groups and an attack goes unclaimed, there may be an additional incentive to free ride and not attack but to claim credit for another’s violence. (p. 426)

When a violent organization carries out an attack, nonviolent groups may have a strong incentive to distance themselves from the act. One common way to accomplish this is to blame the attack on a rival organization. Of course, the organization falsely being blamed for the attack may also not wish to be associated with the stigma of violence. Out of such an environment, a hot-potato problem may arise where groups pass blame to rival organizations in order to avoid the stigmatization of being associated with violence. The hot-potato problem can also arise when a group perpetrates an attack and initially takes credit for it but then retracts and possibly blames a rival group, generally due to backlash or other negative effects of having claimed credit…

The lie of omission is most common; the perpetrator neither claims credit for the attack nor blames the attack on another group. (p. 427)

When civilians are killed in a terrorist attack, the target population will likely be upset and may react in ways that are counterproductive to the perpetrator’s goal(s). Killing civilians can also result in international outcry against the responsible organization. Civilian casualties will often make governments less cooperative with responsible terrorist groups, and governments may react by changing policies that the terrorists enjoyed. For all of these reasons, claiming terrorism can be a poor strategy. (p. 429)

Tanisha M. Fazal, “Dead Wrong?: Battle Deaths, Military Medicine, and Exaggerated Reports of War’s Demise,” International Security, 39:1, Summer 2014, pp. 95-125.

Major advances in military medicine have made battle deaths less likely and nonfatal battle casualties more likely over the past several centuries, particularly since 1946—the same time period that Goldstein, Pinker, and others examine to support their conclusion that war is on the decline. Because these datasets identify wars based on a battle-death threshold that is constant over the entire time period they cover, any apparent decline in war means that war has become less fatal. This observation does not necessarily mean, however, that war has become less frequent… (p. 96)

War, in other words, has become less lethal. The same soldier fighting in the same country’s military today as compared with 200 years ago is much more likely to survive any given military campaign because of improvements in military medicine. To use a constant number for battle fatalities as a measure of the war proneness of society over a 200-year period, therefore, obscures improvements in military medicine and distorts the view of the incidence of war. (p. 113)

The data suggest that both battle deaths and battle casualties are on the decline….From 1946 to 2008, the number of battle deaths declined by 50 percent. By contrast, estimated battle casualties during the same period declined by less than 20 percent. Thus, the data suggest that battle deaths declined more than twice as quickly than battle casualties during this time frame. If this is true, then it may be premature to declare victory on war. (p. 116)

Counting war fatalities as a means to measure the frequency of war is an apparently sensible but, ultimately, flawed strategy given recent developments in technology and warfare. (p. 122)

Rethinking the approach to how scholars of international relations count wars may require significant revisions to major datasets such as the Correlates of War and the UCDP lists of armed conflicts. Both datasets have a fatality threshold for inclusion, and that threshold is constant for the entire period they cover. But if, because of improvements in military medicine, the same conflict that produced 1,200 fatalities in 1860 is likely to have produced 800 fatalities in 1980 (or 35 versus 20, in the case of UCDP), the ahistorical nature of the battle death threshold means that the latter conflict(s) will have been left out of the data set…

The possibility that wars and armed conflicts have been undercounted because existing datasets use a battle death threshold that does not change over time raises questions about the comparability of cases over time and the soundness of quantitative analyses that cover the nineteenth and twentieth centuries, in particular. (p. 123)