Elsevier

Social Science Research

Volume 110, February 2023, 102849
Social Science Research

Where the bombs fell: Measuring compliance with humanitarian treaties

https://doi.org/10.1016/j.ssresearch.2023.102849Get rights and content

Abstract

Compliance with international treaties is the bedrock of the international order. When it comes to international humanitarian treaties, which regulate war-making, the issue of compliance gains urgency as people's lives are on the line. At the same time, measuring states' actions during an armed conflict is known to be exceedingly challenging. Current measures for states' compliance with their international obligations during armed conflict has been incomplete, offering a broad generalization that does not necessarily capture reality on the ground or alternatively based on proxy measurements, which produce a distorted portray of events in relation to obligations. This study suggests geospatial analysis as a gauging tool for states' compliance with international treaties during armed conflict. Examining the 2014 Gaza War as an instrumental case study, this paper underscores the efficacy of this measure and informs current debates on the success of humanitarian treaties and variation in compliance.

Introduction

Many expectations have been placed on humanitarian treaties' ability to influence policy and states' behavior for the better in the context of human rights and armed conflict (Garcia 2015; Hathaway 2007; Rutherford 2000; Simmons 2009). However, the success of humanitarian treaties hinges on our capacity to effectively measure states' compliance. Currently, our compliance measurements of humanitarian treaties during armed conflict rest on textual sources and secondary sources (Morrow 2007; 2008), which vary in quality (Höglund and Öberg 2011). Alternatively, we use quantitative proxies as measurements, such as the volume of civilian casualties (Valentino et al., 2006; Stanton 2016). Those are inconsistent and do not successfully represent compliance given they can be the result of an error or occur inadvertently and within legal parameters. The main obstacle in that regard is our ability to effectively measure compliance or violation in a systematic, standardized matter under conditions of shortage of reliable information, secrecy, and even deception. This paper suggests geospatial analysis as a novel approach for the measurement of states' appliance of violence during an armed conflict. We suggest that this approach utilizes advancement in geo-mapping to create a modular measurement method that avoids the pitfalls of existing solutions and allows a more nuanced assessment of violations. To assess states' compliance with this tool, we examine the locations that were hit with bombs and missiles. By focusing on where the bombs fell, we can improve our assessment of actors' behavior and therefore the degree of states’ compliance.

To determine compliance with humanitarian treaties, this paper examines the Israeli target selection during the 2014 Gaza War as an illustrative case study. As a state with complex compliance record, the focus on Israel helps avoid the selection bias concerns often raised about studying treaty compliance (von Stein, 2005). The 2014 Gaza War epitomizes the most frequent type of armed conflict today—a state combating a violent non-state actor (VNSA). VNSAs introduce major challenges to compliance. They are not signatory members to humanitarian treaties and there is not an effective regulatory mechanism to penalize them. This issue becomes urgent since compliance is associated with reciprocity (Morrow 2007). This study uses the descriptive statistics of a geospatial analysis to examine the Israeli adherence and violation of its humanitarian treaties during the conflict. It compares the ratio of missile/shells hits in predetermined radii around sensitive protected by treaty facilities, ranking the safest across models and measures. This allows a more nuanced and focused examination of compliance than existing measures that at best will focus on casualty count. All facilities examined are considered sensitive for normative reasons and recognized as civilian facilities. They are also protected by international treaties Israel is a signatory member to. At the same time, some of those same facilities were used by Palestinian militants as firing positions and command headquarters (Booth 2014; Chandler 2014; High Level Military Group 2015; State of Israel 2015), complicating their legal status. Military usage of civilian facilities weakens their treaty protection, inviting a justified and proportional retaliation (Jaworski 2003). The result is a hard to measure environment.

This line of inquiry addresses a lacuna in our assessment of states’ compliance with the treaties that regulate war. Measuring compliance is not an unambiguous task, namely, examining whether a state follows the letter of the law or not (Traven and Holmes 2020). Conditions of war present significant challenges for data collection and measurement, making it harder to estimate if a violation took place and who is responsible for it (Wiener 2004). Moreover, given the potential repercussions of a violation, which can include embarrassment in international forums and sanctions (Vennesson 2014), signatory members push back on critique, resist to international pressure, or being evasive in their response (Wiener 2004; Buzas 2017; Dixon 2017). Scholars approached this issue by examining observable behavior and its effects. Morrow and Jo. (2006) and Morrow (2007) employed historical work and journalistic sources to create a novel measure of compliance. Valentino et al. (2006) and Stanton (2016) focused on civilian casualties as a proxy for compliance. However, textual sources do not offer a systematic and standardized measure, and civilian casualties are an important proxy yet presents meaningful biases.1 The gap in measurement makes it harder to understand why states decide to derogate from treaties or to keep them (Hafner-Burton et al., 2011), and if treaties work or not (Valentino et al., 2006; Stanton 2016). It also projects on the discussion of variation in compliance, where states adherence to some treaties is inconsistent (Morrow 2007; 2008). To tackle this gap, we suggest that geospatial analysis can offer a more accurate, nuanced, and standardized measure.

The study of humanitarian treaties' success in influencing policy and state behavior (Cole 2005, 2009; Cottrell 2009; Garcia 2015; Simmons 2009; Wotipka and Tsutsui 2008) can be contextualized in a broader framework and discussion of the increasing role of the International Human Rights Regime in global politics. The human rights regime is an effort piloted by the United Nations to codify human rights in a universally recognized regime of treaties, institutions, and norms. Parallel to this system are the international humanitarian treaties (IHT) that have similar and overlapping concerns as it governs the laws of war (Meron 2000). Scholarly examination of the IHTs' success is regularly associated with controlling the proliferation of weapons of mass destruction (Fuhrmann and Lupu, 2016; Moxley et al., 2010; Price et al., 1996) and limiting types of deadly armament (Price 1998; Rutherford 2000). While successful on some fronts IHTs come short on others. Multiple studies stress the need for better compliance, especially when it comes to Jus in Bello, the laws regulating countries’ conduct during war (Jenks and Acquaviva, 2014; Kelley 2013; Pellandini 2014; Stutz et al., 2015).

Quantitative empirical examination of compliance to IHTs displays mixed results. Valentino et al. (2006) examination of the efficacy of international laws in protecting civilian populations in times of war shows that for the most part treaties fail with providing protection. Their findings indicate that IHTs do not provide strong protection for civilians in times of war. These conclusive results were slightly modified in Morrow's (2007) research on a range of IHTs' effect on states' behavior, highlighting variation in compliance. Morrow presents a four-point degree of compliance model, measuring violations' magnitude, frequency, centralization, and clarity. He shows that while treatment of prisoners of war and the safety of noncombatants displays worse compliance records, laws related to ceasefires, conduct on the high seas, and aerial bombardment are observed.

One of the main indicators for state behavior in studies on compliance with international treaties is the regime type—democratic versus non-democratic/dictatorship. In her book, Simmons (2009) underscores the importance of democratization in the emergence of the human rights regime and the differences between more and less democratic countries in their levels of compliance. Chyzh (2014) examines questions of trust and compliance with international treaties of authoritarian regimes in comparison to democratic ones. Her research specifies that authoritarian regimes are more likely than democracies to enter into international agreements; however, they are less likely to comply with the agreements they sign.

Similarly, in the context of IHTs, the regime type has been a key predictor of compliance. The common assumption about the regime type has been that since democracies are less belligerent (Bremer 1992; Maoz and Russett 1993), and care more about human rights (Prorok and Appel 2014), they would be more likely to comply with humanitarian law. Valentino et al. (2006) challenged this assumption, arguing that regime type fails to prove a reliable predictor of patterns of civilian targeting. Nonetheless, in a later study Morrow (2007) finds mixed trends. Comparing treaty ratification between democracies and non-democracies, Morrow shows that non-democracies’ ratification does not affect behavior while democracies’ ratification signals intent to abide by the treaty standard.

Yet, even for democracies, compliance is not guaranteed and can vary, as demonstrated in Morrow's (2007) degrees of compliance. A glance at the U.S. military operation in Afghanistan exemplifies this variation. Focusing on U.S. military strategies in Afghanistan, Suhrke (2015) study the military's attempts to protect civilians and its failures in the context of aerial bombardment. According to Suhrke, the U.S. military obligation to humanitarian law, in terms of prioritizing it over other elements, shifts in line with the military strategy. Suhrke shows that humanitarian law is prioritized when the military strategy integrates both military and political goals. This prioritization results in a decrease in the number of civilian causalities. Suhrke concludes that the U.S. forces in Afghanistan were rarely responsible for deliberate killings, yet the implications of different prioritization led to multiple casualties, injuries, and displacement. Given the importance and impact of prioritization on humanitarian outcome, which are not well measured with crude categorization such as democratic or not, those findings suggest that we need a more nuanced measurement for compliance, one that will account for prioritization and can better explain variation.

This debate and trends presented rest on the question of measurement; specifically, how do we measure compliance? Different measures will produce different outcomes. Limited measures will not clearly identify actual trends, or otherwise, produce skewed assessments. In other words, we need to know what is measured to be able to relate to the trends presented and to what they indeed represent.

Measuring conflict indicators is notoriously complicated. A popular saying suggests that the truth is the first casualty of war, capturing the challenges of data collection in conflict environments. Those go beyond the expected challenges of data collection and interpretation (Fazal 2014). Actors' incentive to control the narrative and portray a favorable or adversarial image can disrupt the quality and focus of the information released. Trauma and confusion among informants, victims, and witnesses may afford a skewed or misinforming portrayal of events on the ground (Höglund and Öberg 2011). Furthermore, data collection attempts may encounter distrust and suspicions from the local population (Cohen and Arieli 2011). Destruction can cause the loss of information and proof of actors' behavior and actions. Third party monitoring is far from being effective, as it suffers from security and access related restrictions (Brück et al., 2016; Höglund and Öberg 2011). Reliance on secondary sources, such as news outlets, government statements, nongovernmental organizations’ reports, and social media presents a set of challenges that are associated with inconsistency in definitions, biases, and political affiliation that can obscure or distort the evidence presented (Dulić 2011; Möller 2011, 2011berg and Sollenberg 2011). Those issues accrue to massive challenges for researchers.

Advances in quantitative research on conflict manufactured multiple datasets and conflict measures that invited IHL scholars to rethink compliance. The most common set of measures utilized by IHL scholars have been those related to civilian casualties. In their study on IHL protection of the civilian population in times of war, Valentino et al. (2006) focused on the killing of noncombatants as a measure for compliance. Prorok and Appel (2014) and later Bussmann and Schneider (2016) used the same measure to explore the effect of third-party's coercion on states' compliance, suggesting an alternative explanation for the mixed results. Suhrke (2015) examined aerial bombardment IHL restrictions during the Afghanistan War, using the measure of civilian casualties. Studying restraint in civil wars, Stanton (2016) elaborated on this measure, looking at forms of violence against civilians. Starting with civilian death data Stanton identified and categorized types of civilian targeting during war by state and non-state actors. Her measurement focused on mass violence and categorized as control-related violence, cleansing-related violence, and terrorism-related violence.

Morrow and Jo. (2006) introduce the first comprehensive quantitative effort to measure compliance with the creation of a database on compliance with the laws of war in 20th-century interstate wars. They concentrated on nine area issues that are grounded in IHL treaties, exploring them across five measures: magnitude of violation, frequency/extent of violations, centralized control of violations, clarity of violations, and quality of data/evidence. In his research on compliance with IHL, Morrow (2007; 2008) utilized the same databases and measures.

The civilian casualty count, whether it is computed by concentrating on the number of noncombatant death (Valentino et al., 2006) or Stanton's examination of forms of civilian casualties (2016) is a problematic measure regarding IHL compliance. Though civilian casualties' counts are useful for understanding the degree to which IHL protects civilians, they are insufficient measures of IHL compliance as a rule. IHL does not stipulate that civilians cannot be killed or that states cannot inadvertently destroy civilian property. Instead, IHL's principle of distinction stipulates that states cannot intentionally target civilians or civilian facilities. For measurement purposes, this suggests that focusing on civilian casualties may be misleading, an unintended result of lawful actions and not a violation. Furthermore, contemporary conflict environments present challenges that affect civilian protection. When civilian facilities are used for military purposes (Corn 2005; Forrest 2006) their IHL protection is weakened by the principle of proportionality. IHL's principle of proportionality holds that unintentional civilian losses (i.e., collateral damage) cannot be excessive in relation to the specific military advantages anticipated. It regulates the tension between military necessity and civilian immunity (Jaworski 2003) and means that IHL protection should not be abused. The principle is not a carte blanche for actors to use unrestraint coercive force. Instead, it aimed at reducing unnecessary suffering even in situations when actors' use of violence is justified and/or required. This principle places confinements on the use of violence, making sure it is limited to definite military advantage and in disproportion with it (Médecins Sans Frontières 2020).

The issue of proportionality takes precedent when the conflict involves actors that are not well regulated by IHL and as such, are more likely to abuse it. VNSAs, such as terrorist organizations, insurgents, guerilla groups, or criminal organizations are involved in most armed conflicts today. A review of ongoing conflicts demonstrates that VNSAs are the principal adversary in over 95% of them.2 Some of these organizations hold and manage a territory, de facto acting as a state. Yet unlike states, international law and treaties give only a few mechanisms to ensure that VNSAs comply with humanitarian norms. IHTs impose obligations on states yet are limited in dealing with VNSAs.

Regardless of this dilemma and problematic situation, the scholarship on these issues is underdeveloped (Malanczuk 2000; Ronen 2013). The legal basis for underpinning the application of IHTs to non-state armed groups was recognized (Kelley 2013; Murray 2015); however, the international systems' aptitude to enforce VNSA's compliance remains toothless and ineffective. The main hurdle in compelling VNSAs to abide by international treaties is that they are not a party to the agreement. Though including VNSAs as a party to the treaty is possible, the international community fears that this act legitimizes rogue actors (Bongard and Somer 2011; Clapham 2015). Consequently, VNSAs still occupy a dead spot in humanitarian law, creating an odd and complex situation for those who deal with them. The lack of VNSA regulation has implications on states' compliance.

VNSAs take advantage of the international community's helplessness, using states' obligations as protective shields against states' violence (Dunlap and Charles, 2016; Rubinstein and Roznai 2011). This abuse of international law is pronounced around the issue of human shields, when VNSAs use civilian hostages to deter security forces' attacks. Operating within the civilian population and relying on it for logistics and intelligence we can expect VNSAs to show a level of assimilation. Nonetheless, the purposeful use of sensitive locations that enjoy special treatment by international treaties, such as schools, hospitals, or religious houses, as barracks or fire positions is a conscious abuse of international law. This strategy persists in various conflict areas and among multiple VNSAs. It also proved very effective in increasing these actors' survivability during armed conflict.

Those practices create a situation that violates the IHL protection and invites states’ retaliation. Given reciprocity of humanitarian treaties is a key factor in predicting compliance (Morrow 2007), this type of rival, which does not comport with humanitarian law, increases the likelihood of state violation. Even if not violating, according to IHL, under those circumstances, states are allowed to use proportional response to counter or terminate the threat. This means that VNSAs abuse of IHL protection can invite an attack and reduce the level of protection offered by treaties.

Conjointly, IHL's outlook of civilian casualties and the principle of proportionality implies that states are permitted to kill civilians in war if civilian casualties are not intentional and/or not excessive in relation to expected military advantages. At the same time, it is important to stress that states are also obligated to abide by the principle of precaution in their attacks, which requires them to take all feasible precautions to minimize harm to civilians and civilian objects. Accordingly, measures of IHL compliance that concentrate on civilian casualty counts are insufficient since states can comply with their IHL obligations even if they kill a large number of civilians as long as civilian casualties are unintended, proportional to expected military advantages, and that the states took viable protections to minimize harm to civilian. It means that casualties count as IHL measure works when we account for the context of each attack, a very challenging task.

Morrow and Jo. (2006) database goes beyond civilian casualty count, offering a more elaborate measure for compliance that specifies several IHL subject issues. Yet, this database presents its own limitations. The measurements in the database relied on secondary sources. Those sources differ in quality of sources and the quality and information. The sources may reflect biases, change over time, lack authenticity, and not be fully representative of the wider population—the type of biases we can expect with secondary sources (Höglund and Öberg 2011). Morrow and Jo acknowledge the variation in the sources’ quality. They rank each source on a four-level scale from “sketchy evidence” to “excellent documentation.” They also recognize that ongoing development in research on war violations and crimes will probably fill out gaps in the database and may change the results. To summarize, we see that existing measures that focus on civilian casualties to measure compliance present limitations. First, states may kill many civilians while not breaking their IHL obligations, making the casualty count meaningless-compliance-wise. Second, reliance on secondary sources varies in quality and may carry political biases, damaging the accuracy of the measure (Höglund and Öberg 2011).

To overcome the challenges and limitations that current measurements present, we suggest using geospatial analysis, focusing on the spatial manifestation of violations. Geospatial analysis gathers, manipulates and displays social data (such as demographics or events) on maps, intersecting the geography and topography with the social dimension.3 It can be used to assess how dangerous a neighborhood is by inputting crime incidents on a map; and tell more elaborate stories if we add demographics and unemployment rates to the same map. Spatial analysis has been extensively used in Geosciences and Environmental Studies, yet its utility was picked up by other disciplines, such as Urban Studies and Engineering. Social scientists followed with anthropologists mapping excavation sites, demographers mapping population growth in counties, political scientists mapping neighborhood voters’ turnouts, and criminologists mapping crime trends in cities. The implications for human rights analysis were also explored. In some cases, violations have a physical footprint that can be identified and recorded on the spatial plain (destroyed facilities) (Sulik and Edwards, 2010), in others, testimonies and accounts can be displayed with geospatial tools (Madden and Ross, 2009). In both cases, the spatial illustration of violation can accentuate hidden patterns such as clusters or trends. Geospatial analysis can enable scholars and practitioners to determine more accurately IHL compliance. For example, geospatial analysis can establish where bombs and missiles were dropped and their proximity to facilities and objects that enjoy IHL protection, such as hospitals. This does not rely on secondary sources or potentially skewed sources. It can also assess the volume of attacks on protected facilities and their immediate surroundings. The assessment of proximity and volume per protected facility can show violations, and when comparing several locations can underscore trends in violations. The closer a bomb is dropped to a protected facility, the more likely it is that the state has deliberately violated (or overextended) the law, regardless of exactly how many civilians were harmed or how much civilian property was destroyed. Thus, this measure offers nuances that current measures are not equipped to address. Potentially, it can even hint on intention. Across multiple cases within a conflict environment, the frequency of proximity and volume of attacks can indicate that there might be intention, or at least systematic negligence. Essentially, with a big enough sample, trends that indicate an abnormal behavior (i.e., significantly higher rates of bombardment of religious institutions compared to other facilities) can suggest this type of behavior is not random.

To illustrate the advantages of a spatial analysis of IHL, we look at the 2014 Gaza War between Israel and Hamas as an instrumental case study. Using this measure allows us to circumvent the pitfalls of the civilian casualties’ measure, which is highly politicized in this context, and to bank on the advances and detailed geo-spatial reporting of destruction in the aftermath of the conflict. While displaying its own unique characteristics, the 2014 Gaza War also represents a broader general global trend in contemporary conflicts of a democracy fighting a VNSA. The overwhelming majority of conflicts occurring in the last couple of decades transpire between states and violent non-state actors (Pettersson and Eck 2018).

Israel's dialogue with IHTs has been thorny. Its occupation of the West Bank and illegal settlements are at odds with IHL expectations. Furthermore, as an active and chronic combatant, it had repeatedly challenged the boundaries of its humanitarian treaties and commitments. Despite those, Israel is ranked as a democracy according to common social science measures, such as Polity IV and the Democracy Index. It has been constantly engaged with humanitarian rights issues in its political, judicial, and public sphere. Israel is a signatory member to 18 humanitarian treaties, many of them ratified in 1995, and 2000. Those overtime were incorporated into its national laws and laws of war codex (see ICRC database). Looking at compliance, its humanitarian track record improves and degrades along with the ebbs and flow of armed violence. This complexity was illustrated in Simmons' (2009) review of the Israeli commitment to the Convention against Torture, where despite a series of violations and security urgencies, Israel committed itself to the treaty. This complexity, which makes it hard to predict Israel's behavior, addresses one of the main hurdles in studying treaty compliance: selection bias (Downs et al., 1996; von Stein, 2005), making Israel an excellent case study for studying this topic.

Among the human rights and humanitarian treaties Israel is a signatory member to are the III Geneva Convention and the three Hague treaties on the Protection of Cultural Property (Final Act on the Protection of Cultural Property, The Hague, 1954; Hague Convention for the Protection of Cultural Property, 1954; Hague Protocol for the Protection of Cultural Property, 1954) (ICRC Database). Those are the humanitarian treaties we will focus on in our analysis (IHL treaty group). The former ensures the protection of people displaying the emblems of medical religious personnel at times of war, identifying them as performing a humanitarian service. This definition includes hospitals, health clinics, and ministry of health facilities where these medical personnel operate. The latter guarantee the protection of places of worship, including mosques (Table 1). Other sensitive facilities (e.g., schools, universities, and UNRWA schools) are covered by the Additional Protocols of the Geneva conventions (I and II) and the Rome Statute of the International Criminal Court; on both Israel is not a party.4 Humanitarian treaties are part of a system of legal protections. The second layer of protection is provided by the Customary International Humanitarian Law (CIHL), which encompasses the mentioned facilities and others. CIHL is the body of rules of public international law that govern states’ conduct during armed conflict. Israel includes part of the CIHL rules and standards in its jurisprudence and its military law codes. The protection of all civilian facilities is covered by The Principle of Distinction between Civilian Objects and Military Objectives and the Principle of Precautions in Attack, which requires the differentiation between civilians and military objects. Mosques specifically have additional protection by those laws with Attacks Against Cultural Property that provides additional protection to religious houses and universities, safeguarding them from attacks. CIHL covers all facilities in our sample, and thus functions as an underline condition and is not included in one of the groups we measure. States also sign treaties and conventions that are not war focused yet can influence the status of facilities and individuals during war. This is the third layer of protection. In the case examined Israel is signed on several of those treaties; one that protects UN facilities, 1946 Convention on the Privileges and Immunities of the UN, in particular Article three of the Convention, that “the premises of the United Nations shall be inviolable.” This includes UNRWA schools, and another one that protects schools and universities, Convention on the Rights of the Child. Those are part of the second group we study (Other treaty group).

CIHL protects all non-military targets, meaning, any attack on a non-military target is a violation and a war crime. Yet, treaties add a normative restriction, hence our focus on them in this study. Signing a treaty that stipulates protection of some facilities (such as hospitals) increases states' perceived commitment. Moreover, while clear in the context of interstate conflict, the boundaries between military and non-military targets blurs in an intrastate conflict, where VNSAs often assimilate with the local population and use civilian facilities. For example, insurgents that use an active civilian hospital as a headquarters will effectively null its CIHL protection, yet it is unclear if states will in fact attack this location due to normative restrictions and treaty obligations. Thus, we focus in our analysis on the layers of treaties, examining their potential association with states’ behavior. By doing so, we attempt to isolate the unique impact of set of treaties, IHL or others.

Israel's rival in the Gaza War, Hamas, undertook full control over the Gaza Strip after its electoral victory and later triumph in the 2006–2007 Palestinian Civil-War (Schanzer 2008). However, the Gaza Strip under Hamas became an insulated island as the Palestinians' immediate and most influential partners, Israel, Egypt, and the U.S., designated Hamas as a terrorist group. The terrorist brand crippled the organization, as it lost access to most international markets and potential partners. On the domestic front, Hamas faced antagonistic relations with the Palestinian leadership in the West Bank that damaged both its legitimacy and ability to rule. This isolation was exacerbated by the Israeli and U.S. sanctions against the Hamas government, the fighting cycles of Hamas and Israel, and the Israeli and Egyptian blockade of the Gaza Strip. Deprived of most of its resources and legitimacy, Hamas utilized one of its few leverages—its ability to produce instability via violence. It was a simple and efficient tactic of propagating fear, mostly by raining rockets on Israeli cities and then pricing the conditions for peace with economic or political concessions. It was a risky tactic, which often cost Hamas and Gazans in life and infrastructure yet was able to yield some of the concessions the besieged organization anticipated. An example of those is Israeli concessions regarding Gaza fishermen after Operation Pillar of Defense, allowing them to fish within the area of six nautical miles of Gaza.

In the summer of 2014, a series of events quickly rolled into a daily exchange of rocket and missile fire and culminated with a wide Israeli military operation in the Gaza Strip. During the conflict, Hamas took advantage of the legal asymmetries in international law that favor VNSAs (Bassiouni 2007; McKeown 2015). It exploited its position of power as sovereign in Gaza (Berti and Gutiérrez 2016) to coerce cooperation from citizens, businesses, and international organizations as it constructed fortifications in or near their facilities (High Level Military Group 2015; State of Israel 2015). Known for their unique status hospitals, schools, religious houses, and UNRWA facilities became not only sanctuary for people escaping war, but also a kind of a shield for Hamas operatives, arms, and military infrastructure against Israeli attacks (Chandler 2014; McCoy 2014). Amidst the conflict, and right after, numerous reports and photos of news outlets and the IDF spokesman stressed Hamas insurgents' practice of launching missiles in the vicinity of these kinds of sensitive installations (Chandler 2014; High Level Military Group 2015; Taylor 2014; State of Israel 2015). This strategy's rationale was to deter Israel from intercepting Hamas' operatives both by using the civilians in these facilities as human shields and by exploiting the protection these buildings are afforded by international law. Notwithstanding its controversiality, this tactic is a common practice among multiple VNSAs in the MESA region today, such as Hezbollah and ISIL, and others across the globe and throughout history (see Hoffman, 2018; Marighella 2021 [1971]; Dolan et al., 2018). The Gaza Strip main hospitals, such as Al-Shifa, turned into “a de facto headquarters for Hamas leaders” (Booth 2014). Furthermore, on three separate instances UNRWA staff discovered rocket caches hidden by Hamas in its facilities (McCoy 2014). Those practices harmed treaty protection of sensitive facilities and opened the door for a proportional response.

The 2014 war was severe, different than the usual skirmishes between the belligerents. During the conflict, Hamas launched about 4500 rockets and mortar shells at Israeli cities. The range of those missiles covered over three-quarters of the entire state of Israel and targeted three of Israel's most populated cities: Jerusalem, Tel Aviv, and Beer-Sheva. The daily average of rockets launched was 86, with the Israeli anti rockets system, Iron Dome, overwhelmed by the constant barrages and intercepting only 735 rockets. 225 of the rockets landed in civilian urban spaces, resulting in 6 civilian deaths and 836 injuries (Shafir 2014). Furthermore, dozens of rockets targeted Ben Gurion Airport, the sole Israeli international commercial airport, setting off a 48-h-long ban of international airlines flights to and from Israel (Mouawad 2014).

Those attack outcomes can be considered as high cost for Israel that may create urgency around response and might even lead to treaty violations. However, how are those attack outcomes any different from earlier cycles of violence? Since the Israeli disengagement from Gaza in 2005, Israel and Hamas have been fixed in a vicious cycle of violence. Each of those violent cycles involved rockets launched to Israeli cities. During that period, Israel broke off its treaty obligations regarding the protection of hospitals and UN facilities, which represent facilities from the IHL treaty and other treaty groups, only in two instances, during the 2008–9 Gaza War and the 2014 Gaza War (Fig. 1).5 Fig. 1 illustrates the correlation between the number of rockets and mortar attacks from Gaza, number of Israeli casualties, and the Israeli targeting hospitals and UN facilities, in divergence from its treaty obligations. On those two wars Israel failed to stop or contain Hamas’ attacks. The two demonstrate the highest rates of rocket attacks and Israeli casualties. Those figures indicate that not every attack leads to treaty breakdown, only those that include massive rocket fire, which represents a high cost for Israel. Furthermore, Fig. 1 shows that at least regarding those two facilities (and related treaties), Israel is not a chronic violator, yet it is also not in full compliance. Its violations are clustered around the two wars.

The Iron Dome failure to defend Israeli cities, along with Hamas' successful attacks on Israeli cities and airport, pressed Israel to adopt the proportionality response and take greater risks in its target selection that ended in the violation of its treaty obligations. Israel decided to target Hamas' missile teams that operated from protected areas while using what Israel assessed as a proportional response. Though within the legal boundaries of IHL in those situations Israel did break its treaty obligations. The success of those violations in suppressing Hamas' fire is unknown. Yet, the unintended consequences of those specific Israel's violations were Palestinian casualties. Israel bombarded UNRWA schools and hospitals, with IDF shells hitting UNRWA facilities harboring refugees seven times, causing 56 deaths and hundreds of injuries (Jalabi 2014).

What those descriptions of events can tell us about Israel's compliance with IHL and its treaty obligations during that conflict? A discussion that focuses on a binary outcome will suggest that Israel violated its treaty obligation and disregarded CIHL; plain and simple. Contrarily, a focus on variation in compliance may suggest that Israel hitting at least some of those facilities were errors occurring while Israel was exercising the proportionality principle. To better understand Israel's compliance, we need better measurements that would account for the complexity of the battleground. Using geospatial analysis of where the Israeli bombs hits, we can calibrate our understanding of compliance while accounting for variation. We can examine the level of protection that civilian facilities that are covered by IHL treaties signed by Israel (IHL treaty group) and other facilities covered by non-IHL treaties (other treaty group) enjoyed during the conflict to explore which of the two influence the state, if any. Because of Israel's treaty commitments, health facilities and mosques (IHL treaty group) enjoy a special status, legal and political, and therefore one can suggest that they will enjoy better protection than others. We would assume that IHL treaty obligations place additional constraints on signatory members and therefore would make facilities protected by them safer compared to civilian facilities protected merely by other treaties. To put it simply, using geospatial measures, we assume that facilities in the IHL treaty group and their immediate surroundings will show fewer hits compared to the other treaty group. Given that a great proportion of the armament used by Israel was guided missiles, meaning it produces significant accuracy in its target selection, we can treat the number of hits and locations as degrees of compliance.

Section snippets

Data and methods

The data is taken from the United Nations Institute for Training and Research (UNITAR) publication: The Gaza Crisis Atlas (Saidane 2014). The atlas geolocate every building, road, or strip of land damaged during the conflict on a map. It categorizes damage into four degrees of severity, ranging from full destruction to a visible crater. We converted this data to GIS format, computed interactions between hits and infrastructures, using the buffer and join options, and exported the new data to

Discussion

This study presents geospatial analysis as a tool to identify and measure compliance with IHL treaties. Traditionally, effective measurement of treaty compliance during an armed conflict has been challenging and limited (Brück et al., 2016; Höglund and Öberg 2011), making it difficult for scholars, legal experts, and policymakers to assess treaties' efficacy and contribution. Examining the 2014 Gaza War we use geospatial analysis of where the bomb fell to probe the relative safety of facilities

Conclusion

IHTs are an attempt to regulate and mitigate the effects of one of the most destructive behaviors in the international system—the appliance of organized violence by states. Indeed, with such progressive goals, compliance with IHTs regularly receives much attention from academics, politicians, leaders, humanitarians, and policy experts. Without compliance, IHTs are merely empty promises that could be misused to garnish leaders in festive celebrations. Unfortunately, compliance with IHTs is not a

References (89)

  • Amnesty

    Nigeria: Satellite Images Show Horrific Scale of Boko Haram Attack on Baga” Amnesty International

    (2015)
  • B'Tselem

    Fatalities during operation cast lead

    B’Tselem – The Israeli Information Center for Human Rights in the Occupied Territories

    (2021)
  • Jutta Bachmann et al.

    Gaza 2014: findings of an independent medical fact-finding mission

    Physicians for Human Rights-Israel

    (2014)
  • M. Bassiouni

    The new wars and the crisis of compliance with the law of armed conflict by non-state actors

    J. Crim. Law Criminol.

    (2007)
  • Benedetta Berti et al.

    Rebel-to-political and back? Hamas as a security provider in Gaza between rebellion, politics and governance

    Democratization

    (2016)
  • Pascal Bongard et al.

    Monitoring armed non-state actor compliance with humanitarian norms: a look at international mechanisms and the Geneva Call Deed of Commitment

    Int. Rev. Red Cross

    (2011)
  • William Booth

    While Israel held its fire, the militant group Hamas did not

  • Stuart A. Bremer

    Dangerous dyads: conditions affecting the likelihood of interstate war, 1816-1965

    J. Conflict Resolut.

    (1992)
  • Tilman Brück et al.

    Measuring violent conflict in micro-level surveys: current practices and methodological challenges

    World Bank Res. Obs.

    (2016)
  • Margit Bussmann et al.

    A porous humanitarian shield: the laws of war, the red cross, and the killing of civilians

    Review of International Organizations

    (2016)
  • Zoltan Buzas

    Evading international law: how agents comply with the letter of the law but violate its purpose

    Eur. J. Int. Relat.

    (2017)
  • Adam Chandler

    Hamas quietly admits it fired rockets from civilian areas

  • Olga Chyzh

    Can you trust a dictator: a strategic model of authoritarian regimes' signing and compliance with international treaties

    Conflict Manag. Peace Sci.

    (2014)
  • Andrew Clapham

    Human rights obligations for non-state-actors: where are we now?" forthcoming in doing peace the rights way: essays in international law and relations in honour of louise arbour

  • Nissim Cohen et al.

    Field research in conflict environments: methodological challenges and snowball sampling

    J. Peace Res.

    (2011)
  • Wade M. Cole

    Sovereignty relinquished? Explaining commitment to the international human rights covenants, 1966-1999

    Am. Socio. Rev.

    (2005)
  • Wade M. Cole

    Hard and soft commitments to human rights treaties, 1966-2000(1)

    Socio. Forum

    (2009)
  • Geoffrey S. Corn

    Snipers in the Minaret-What Is the Rule-The Law of War and the Protection of Cultural Property: A Complex Equation

    (2005)
  • M. Patrick Cottrell

    Legitimacy and institutional replacement: the convention on certain conventional weapons and the emergence of the mine ban treaty

    Int. Organ.

    (2009)
  • Stephen D. Davis

    Controlled warfare: how directed-energy weapons will enable the US Military to fight effectively in an urban environment while minimizing collateral damage

    Small Wars Insur.

    (2015)
  • De Smith et al.

    Geospatial analysis: a comprehensive guide to principles, techniques and software tools

    (2007)
  • Jennifer Dixon

    Rhetorical adaptation and resistance to international norms

    Perspect. Polit.

    (2017)
  • Thomas M. Dolan et al.

    Where the insurgents aren't: rurality, information, and nonterritorial insurgency

    J. Conflict Resolut.

    (2018)
  • George W. Downs et al.

    Is the good news about compliance good news about cooperation?

    Int. Organ.

    (1996)
  • Tomislav Dulić

    Peace research and source criticism

    Understanding Peace Research: Methods and challenges

    (2011)
  • Dunlap et al.

    No good options against ISIS barbarism? Human shields in 21st century conflicts

    AJIL Unbound

    (2016)
  • Tanisha M. Fazal

    Dead wrong?: battle deaths, military medicine, and exaggerated reports of war's demise

    Int. Secur.

    (2014)
  • Craig JS. Forrest

    The doctrine of military necessity and the protection of cultural property during armed conflicts

    Cal. W. Int'l LJ

    (2006)
  • Matthew Fuhrmann et al.

    Do arms control treaties work? Assessing the effectiveness of the nuclear nonproliferation treaty

    Int. Stud. Q.

    (2016)
  • Denise Garcia

    Humanitarian security regimes

    Int. Aff.

    (2015)
  • Paul G. Gillespie

    Weapons of Choice: the Development of Precision Guided Munitions

    (2006)
  • Hafner-Burton et al.

    Emergency and escape: explaining derogations from human rights treaties

    Int. Organ.

    (2011)
  • Oona A. Hathaway

    Why do countries commit to human rights treaties?

    J. Conflict Resolut.

    (2007)
  • Darren Hawkins et al.

    Partial compliance: a comparison of the European and inter-American courts of human rights

    J. Int'l L & Int'l Rel.

    (2010)
  • An Assessment of the 2014 Gaza Conflict

    (2015)
  • Bruce Hoffman

    Inside Terrorism

    (2018)
  • Kristine Höglund et al.

    Doing empirical peace research

    Methods and challenges

    (2011)
  • Human Rights Council. 2019. Report of the detailed findings of the independent international Commission of inquiry on...
  • Raya Jalabi et al.

    Gaza crisis: a closer look at Israeli strikes on UNRWA schools

  • Eric Jaworski

    Military necessity and civilian immunity: where is the balance

    Chinese J. Int'l L.

    (2003)
  • Chris Jenks et al.

    Debate: the role of international criminal justice in fostering compliance with international humanitarian law

    Int. Rev. Red Cross

    (2014)
  • Morgan Kelley

    Challenges to Compliance with International Humanitarian Law in the Context of Contemporary Warfare

    (2013)
  • Nicolas Lamp

    Conceptions of war and paradigms of compliance: the ‘new war’challenge to international humanitarian law

    J. Conflict Secur. Law

    (2011)
  • Marguerite Madden et al.

    Genocide and GIScience: integrating personal narratives and geographic information science to study human rights

    Prof. Geogr.

    (2009)
  • Cited by (0)

    View full text