By 2019, Mali’s civil war had dragged on for seven years, as rebel groups engaged in violence for control of the state. The epicentre was the Mopti district, which experienced more intense hostilities than other regions. It was also the deadliest district for civilians, as rebels targeted them directly. Violence earlier in the decade motivated the deployment of the United Nations Multidimensional Integrated Stabilization Mission (MINUSMA), a peacekeeping operation (PKO) that, by 2019, had become one of the largest in the world. Yet, MINUSMA’s substantial capacity was unable to mitigate the violence and begin the transition to peace. Importantly, the operation was failing to slow an escalation in the late 2010s that was increasingly defined by attacks on civilians (Human Rights Watch 2018). Confidence in the operation waned, and questions arose as to continued civilian support for its deployment (Smith Reference Smith2017).
To confront growing violence and limit rebel attacks, MINUSMA invested in aid projects in coordination with its peacekeeping activities. In one such effort, MINUSMA initiated a Quick Impact Project (QIP) to build a radio tower to project transmissions beyond Mopti town to more rural areas of the district. QIPs are aid investments that are ‘designed to build confidence in the mission, its mandate or the peace process’ in response to targeted community needs (United Nations Reference Harvey, Cd and Fearnley2012, 224–225). Mamadou Bocoum, the Regional Coordinator of Radio and Television Broadcasting Union in Mopti described the investment as critical to reducing violence there. Referencing the importance of the radio’s reach to the ongoing interethnic violence that led to death and displacement in the region, Bocoum stated, ‘We broadcast a variety of messages promoting peace, nonviolence and healing. … Getting timely, effective, fair and accurate information to every village, every community, every citizen is critical in keeping a lid on tensions, particularly now with an upsurge of violence in the area. …The transmitter provided by MINUSMA allows us to do that’ (United Nations 2019).
The transmitter project was one of twenty QIPs that were ongoing in Mopti in 2019, totalling nearly $5 million. This was a substantial increase over previous years and a 50 per cent increase over 2018’s peak. Attacks on civilians concurrently fell by 75 per cent from 2018 to 2019.Footnote 1 At the same time, MINUSMA’s hard power in the form of military personnel and material, shown in recent work to be consequential to civilian protection, remained essentially constant during this time. In fact, the number of blue helmets deployed to Mali declined from late 2018 and early 2019 through the end of the year (Kathman Reference Kathman2013).
This raises the question of whether the UN can use targeted aid effectively as part of its peacekeeping efforts to reduce violence in civil war. This is especially interesting, as the literature on aid provision in civil war contexts has not produced consistent results with regard to the effect of aid on violence intensity. Indeed, recent studies have indicated that some characteristics of aid can have deleterious effects on the well-being of civilians and the course of conflict, while others can be beneficial to civilian safety and the peace process. At the same time, the literature on UN peacekeeping has indicated that UN PKOs can be effective in confronting violence. As part of these efforts, the UN has increasingly committed to QIPs as an element of pursuing successful peacekeeping outcomes (United Nations Reference Harvey, Cd and Fearnley2012). Yet, we know little about whether QIPs work in support of or at cross purposes with peacekeeping’s general effectiveness.
We argue that for QIPs to be an effective element of UN efforts toward peace and civilian protection, the UN and its aid investments must do two things. First, they must disincentivize rebels from exploiting aid through violence as part of rebel efforts to improve their standing relative to their opponents. Looting aid projects through violence and exploiting these resources to cheaply expand rebel capacity are risks associated with aid investments to conflict zones; therefore, QIPs must have characteristics that avoid these dangers. Second, to be effective, QIPs must incentivize rebels to reduce or avoid violence and aid exploitation out of concern for disrupting civilian loyalties in support of their movement. Rebels often rely on civilian support to remain viable in their challenge to the government. Thus, QIPs must have characteristics that risk reducing civilian support for rebels if the aid projects are exploited by them.
On the first issue, the UN is generally viewed as relatively impartial and does not seek to manipulate the strategic balance between conflict factions. Further, even if rebels wish to exploit aid to augment their relative standing, QIPs are not easily lootable and fungible to use in war, making it difficult for rebels to exploit QIPs to improve their relative fighting capacity. This has the effect of reducing the value of violence, thereby disincentivizing rebels from engaging in it. On the second issue, we show that QIPs are directly negotiated with local communities. Rebels, desirous of civilian support, are thereby incentivized to engage cooperatively with the population since violence would disrupt the aid that civilians themselves have sought to acquire.
We test the relationship between QIPs and rebel attacks on civilians with data from the Central African Republic (CAR), the Democratic Republic of the Congo (DRC), Mali, and South Sudan. We aggregate geo-located data on QIPs and rebel violence to the district-month and find that increasing QIP investments are associated with a decrease in rebel attacks on civilians.
UN Peacekeeping Effectiveness, Operational Objectives, and Peacekeeping Power
The recent literature on UN peacekeeping in civil war finds that peacekeeping is generally effective in its efforts to limit violence and pursue peace (Walter et al. Reference Walter, Howard and Fortna2021). The theorized mechanisms are several for this ameliorative effect. Peacekeeping helps to resolve the security dilemma and the associated commitment problems that beguile combatants’ efforts at peace (Walter Reference Walter2002). PKOs can improve the sharing of information between parties that resolves information asymmetries, builds trust, and reveals the rising cost of violence for the purpose of achieving political objectives (Hultman et al. Reference Hultman, Kathman and Shannon2019).
Further, peacekeeping creates space for negotiations and conflict resolution processes. By seeking the consent of the factions, PKOs confer legitimacy upon them. The UN is also the lone representative of the global community. This improves the prospects that recognized combatant groups have a stake in the country’s future, which is critical to compliance with peace processes (Jo Reference Jo2015; Fazal Reference Fazal2017). Impartiality is a core principle of peacekeeping that the UN reiterates with the approval of each operation (Tull Reference Tull2018). The UN’s impartial application of mandated tasks toward achieving peace does not fundamentally favour a particular side in the conflict, short of the factions’ own efforts to obstruct or thwart a PKO’s mandate implementation (Benson and Kathman Reference Benson and Kathman2014). The UN is thus unlike other third parties, giving it unique influence over conflict and peace processes (Fortna and Howard Reference Fortna and Howard2008; Walter et al. Reference Walter, Howard and Fortna2021).
Central to its objectives in the pursuit of peace are the UN’s efforts to reduce violence. While combat hostility is often considered a driver of various forms of violence, not all PKOs are mandated to confront battlefield contests directly. However, all recent PKOs have been mandated to protect civilians from violence. Since its inception, but particularly since the renewed effort to identify means of improving peacekeeping success that came with the publication of the Brahimi Report (United Nations 2000), UN peacekeeping has focused its efforts on protecting civilian populations from violence in war contexts. The protection of civilians (PoC) mandate is unlike any other peacekeeping task in its routine inclusion in PKO mandates. Further, civilian protection is a core peacekeeping value that defines operational success and failure. As former Secretary-General Dag Hammarskjöld claimed, even when operating only under a self-defence mandate with no marching orders to protect civilians, blue helmets could not justify standing by when civilians suffer atrocities (see Wills Reference Wills2009, 267–272).Footnote 2
Much of the research on peacekeeping effectiveness in protecting civilians has focused on the UN’s ability to project hard power. To resolve the security dilemma, PKOs separate factions and safeguard civilians. To resolve information asymmetries, peacekeepers monitor combatant activities and patrol frontlines. These activities rely upon hard power, including militarized tools such as armed personnel and combat materiel. More troops, police, and transport equipment allow for larger buffer zones, greater demobilization, and additional patrols. Consistent with this, empirical assessments reveal a positive effect of hard power on the ability of UN PKOs to protect civilians (Hultman et al. Reference Hultman, Kathman and Shannon2013; Bove and Ruggeri Reference Bove and Ruggeri2016; Kjeksrud Reference Kjeksrud2023).Footnote 3
Yet, hard power is just one class of tools at the UN’s disposal. Indeed, the UN has increasingly invested in the projection of soft, ‘non-militarized’ power.Footnote 4 As Howard (Reference Howard2019) argues, PKOs influence conflict through the exercise of hard and soft power. The UN seeks to coerce with its military capacity short of the use of offensive force, persuade groups on the UN’s path to peace, and induce cooperation with financial incentives. While most research on civilian protection appears to focus on the means by which the UN wields its hard power in pursuit of peacekeeping goals through coercion, Howard shows that this is just one class of power projection, and soft power plays an important role.
The UN’s Quick Impact Projects serve as an important tool in these additional means of influence, seeking to persuade via financial inducement. QIPs do this by providing financial investments in public projects that are preferred by the communities to which these resources are deployed. The UN’s peacekeeping hard power has generally been shown to be effective in reducing violence by rebel groups. Given the government consent necessary for operational activities, research has found UN peacekeeping to be an effective tool primarily for addressing rebel group violence (Fjelde et al. Reference Fjelde, Hultman and Nilsson2019).
Yet, it is these very groups for whom foreign aid can motivate pillaging, exploitation, and attacks on civilians (Wood and Sullivan Reference Wood and Sullivan2015). Rebels face a power asymmetry in their contest with the government, and coercive resource acquisition from the population and foreign aid investments is an attractive option for these groups to improve their standing. Victimization is the means rebels use to plunder and exploit these resources, which stands at cross-purposes to UN PKO objectives. It is thus reasonable to assess whether QIPs serve to support operational objectives to impede violence against civilians or only exacerbate the conditions that drive violence by rebels, thereby undermining confidence in PKOs and leading to operational failure.
Quick Impact Projects and Pursuit of Peacekeeping Objectives
QIPs are a special form of financial investment in PKO host states. Projects are narrowly targeted to support the ongoing operation. By building local support, the UN hopes to generate goodwill for continued peacekeeping while building confidence amongst local actors that peace efforts will ultimately succeed. QIPs are meant as short-term, low-cost/high-return, high-visibility projects. By initiating a QIP, the UN’s civil affairs personnel attached to a PKO engage in consultation with local residents, stakeholders, and elites. From these consultations, investments in particular projects are determined by the recipient community’s preferences. The projects are then carried out in coordination with local authorities and nongovernmental organizations (United Nations Reference Harvey, Cd and Fearnley2012).
Projects include development assistance and infrastructure improvements, like the construction and repair of roads and bridges (Mbugo Reference Mbugo2022). Other projects focus on social impacts, such as building health clinics (Mahamidou and Kristianto Reference Mahamidou and Kristianto2020), constructing or refurbishing schools (Chuol Reference Chuol2017), or improvements to valued cultural and community sites (Kinzli Reference Kinzli2018). Similarly, QIPs can enhance local security through investments in security infrastructure (Andersson Reference Andersson2018) and access to information (United Nations 2019).
However, QIPs are not often associated with the more recognizable security activities carried out by armed peacekeepers, like battlefield intercession, disarmament, and patrols. QIPs do not endeavour to reduce the killing capacity of wars. Instead, QIPs include projects that seek to improve conflict resolution by motivating more cooperative behaviour by otherwise violent actors. They seek to avert conditions that lead to hostilities. For instance, QIPs have sought to improve roads to allow civilians to escape violence (Mapendano Reference Mapendano2021). Investments have offered job training to inmates and improved prison conditions, as these investments limit prospects for future violence by those arrested (United Nations 2017; Phumisa Reference Phumisa2023). These projects attempt to improve general security and reduce the fragility of conflict environments that make these contexts prone to spirals of violence.
QIPs and Civilian Protection
Given the goal of QIPs to mitigate insecurity, meet local community needs, and build trust in the peace process, we can assess their effectiveness in terms of supporting basic peacekeeping goals. The most basic of these is civilian protection. Below, we consider the theoretical link between QIPs invested during civil war and their consequences for rebel use of civilian victimization.
Rebel Motivations for Victimization and UN Aid
To understand the effect of QIPs on rebel use of civilian victimization, we first consider rebel motivations for employing such violent tactics. While this is a rich literature, much research suggests that rebels use victimization instrumentally. Rebels target civilians purposefully. Violence against civilians is not simply an unfortunate by-product of war. Rather, rebels use this violence to achieve their war aims. Anti-civilian violence can be used to shape civilian loyalties in contested territories (Kalyvas Reference Kalyvas2006), overcome recruitment problems (Kalyvas and Kocher Reference Kalyvas and Kocher2007), cow the population (Balcells Reference Balcells2010), and obtain fighting provisions (Wood Reference Wood2010; Hultman Reference Hultman2007).Footnote 5
Central to these instrumentalist arguments is that factions can employ anti-civilian violence as a cost-effective means to acquire resources. For rebels, this resource benefit comes in two basic forms. The first is the direct fighting capacity benefits that can result from civilian victimization: violence can be used to acquire relevant warfighting goods. This can include coercing recruits, securing shelter, and accessing arms. In other words, victimization can be used to mobilize greater capacity to be wielded on the battlefield.
Second, combatants can use victimization to shape civilian behaviour. Rebels tend to be substantially weaker than their regime opponents. To survive and effectively challenge the regime, rebels require the population’s support. Yet, civilians are apolitical to the extent that their primary motivation in conflict is to survive over supporting a particular faction. As such, the population is disposed to supporting the stronger faction in war (Wood Reference Wood2010). This is often a disadvantage to rebels who can thus be motivated to use violence to separate civilians from the regime, acquire information from them about government forces, and motivate their withholding of information from the regime about rebel locations and operations. Providing goods and services to the population is another less violent means by which to secure loyalties. However, benefits in the form of security and other public goods are very costly, especially for rebels, given their resource deficiency. As a result, rebels can use violence to shape civilian loyalties in their favour in a cost-effective manner (Kalyvas Reference Kalyvas2006).
Since QIPs are resource investments, a concern is that such an infusion can motivate rebel violence for the purpose of capture in order to build their conflict capacity or to shape civilian support. Since QIPs serve to expand the resource pie, achieving peacekeeping goals can be difficult when projects attract predation and rent-seeking (Grossman Reference Grossman1992; Addison et al. Reference Addison, Le Billon and Murshed2002; Nielsen et al. Reference Nielsen, Findley, Davis, Candland and Nielson2011). Rebels may also seek to exploit UN investments to fund rebellion or claim credit for projects to purchase civilian allegiances (Jackson Reference Jackson2018; Breslawski Reference Breslawski2023). QIPs might then work at cross purposes to the UN’s intent, exacerbating war (de Wall Reference de Wall1997; Maren Reference Maren2009).
For QIPs to effectively reduce rebel use of victimization in line with the UN’s operational objectives, these projects must successfully achieve two goals. First, QIPs must disincentivize rebels from exploiting projects as part of their interest in expanding their fighting capacity. Resource infusions are associated with less violence to the extent these resources are less fungible to warfighting through looting and victimization (Wood and Sullivan Reference Wood and Sullivan2015; Wood and Molfino Reference Wood and Molfino2016; Narang Reference Narang2015). Further, resource commitments seen as favoring one side can motivate violence for capacity expansion (Narang and Stanton Reference Narang and Stanton2017). Thus, to disincentivize rebel victimization, QIPs must avoid harming the standing of rebels relative to their opponents.
Second, QIPs must incentivize rebels to reduce or avoid violence and aid exploitation out of concern for losing civilian support. When civilians benefit from resource commitments, the costs to rebel groups for exploitation increase (Berman et al. Reference Berman, Felter, Shapiro and Troland2013). QIPs must directly benefit the population, resulting in fewer benefits to rebels for plundering it. Doing so should threaten the rebel group’s ability to secure civilian support. If QIPs are targeted to local needs, rebel violence in exploitation efforts should be accompanied by the danger of withheld civilian allegiances. Below, we theorize on QIPs’ ability to pursue these two objectives.
Disincentivizing Victimization for Capacity Expansion
Resource infusions into war zones are less likely to motivate victimization when aid is provided without threatening the relative fighting capacity of the factions. The impartiality of aid delivery is thus a centrally important factor disincentivizing victimization. QIPs are deployed by the UN, which is generally recognized by combatants and civilians alike as acting on behalf of the global community. UN aid is thus not imbued with the geopolitical interests that are inherent elements of intervention and aid from other parties, which research has shown to yield aid ineffectiveness (Dreher et al. Reference Dreher, Eichenauer and Gehring2016).Footnote 6
UN QIPs are not expressions of geopolitical interests, political preference for a particular party, or bias in favour of a faction’s war effort.Footnote 7 As such, aid from the UN should be considered less threatening to rebel groups’ relative power standing. The UN is also impartial toward conflict actors, short of a preference for commitment to the peace process. In other words, the UN generally does not intervene in support of one faction’s effort to win. Rather, the UN pursues resolution, and it is generally recognized by the combatants as being impartial toward these ends. While PKOs may not be neutral in terms of applauding parties that commit to peaceful practices relative to those that disrupt the peace process, the fact that PKOs do not seek to manipulate civilian support for or against a particular group makes them unlike other third-party interveners.Footnote 8
Thus, QIPs should not be perceived by rebel groups as being invested to tip the strategic balance. This stands in contrast to interventions by many other actors (Regan Reference Regan2002; Balch-Lindsay et al. Reference Balch-Lindsay, Enterline and Joyce2008; Thyne Reference Thyne2009). Resource infusions that do not disrupt the balance of capabilities are less likely to instigate victimization (Wood Reference Wood2010). As such, QIPs should be nonthreatening to rebels, thereby reducing their incentives to sever with violence the civilians’ access to the benefits associated with QIPs.
A second means by which resource investments disincentivize violence against civilians is by reducing the incentives for predation and plunder. Shelter, medical supplies, and foodstuffs have clear benefits to civilian populations in war-torn regions. However, these types of goods are also directly fungible to a rebel group’s warfighting capacity. Rebels can transform these goods into improved military performance through sheltering, treating, and feeding their forces. This fungibility creates rebel incentives for violence and plunder (Findley et al. Reference Findley, Powell, Strandow and Tanner2011).
Yet, QIPs are less fungible to rebel fighting capacity and thus cannot be easily extracted via violence to augment their power. This disincentivizes rebel violence for exploiting QIPs. When a QIP investment refurbishes a cultural meeting place, supplies desks to a school, or improves conditions in a prison, these materials are not easily translated to combat. For example, the UN Mission in South Sudan implemented a QIP to build a solar-powered lighting system in Juba. The goal was to improve lighting in areas around a hospital to reduce crime and attacks on civilians, specifically to limit attacks on hospital workers to improve patient care (Andersson Reference Andersson2018). Projects of this sort are not easily lootable, and there is little that is fungible to conflict.Footnote 9
QIPs implemented by UN operations thus have features that disincentivize exploitation by rebel groups. They are implemented in a way that limits changes to the status quo power balance, thereby reducing incentives for violent extraction. QIPs are also limited in their fungibility to use in war, thus decreasing the usefulness of rebel violence against civilians to gain resource access. As such, rebels are disincentivized from using violence to expand their capacity.
Incentivizing Violence Reduction for Civilian Support
QIPs are determined in consultation between the PKO and the local community slated to receive the project (United Nations Reference Harvey, Cd and Fearnley2012, Chapter 12). Research on peacekeeping effectiveness shows that operations are successful and avoid hazards when they engage with local populations (Otto Reference Otto2019). Autesserre (Reference Autesserre2010) shows how multidimensional mandates require civilian participation. Participation generates support from the population, which manufactures support for the peace process. Community consultation over aid investment also improves effectiveness in meeting aid objectives (Berman et al. Reference Berman, Felter, Shapiro and Troland2013, Sexton and Zürcher Reference Sexton and Zürcher2024). Top-down foreign aid can be divorced from community needs. If not decided through consultations, aid may not map well onto the local context. As such, aid can be more susceptible to predation by rebels. When the benefits of aid to the population are limited, the downside risk to rebels for depriving civilians of these benefits is not as significant (Breslawski Reference Breslawski2022). This exploitation is often achieved through violence, as rebels benefit from looting in pursuit of war aims (Wood and Sullivan Reference Wood and Sullivan2015).
QIPs are unique in that they are implemented after comprehensive consultations with community members, elites, and local stakeholders. These consultations consider implementation plans like locations to which aid should be directed, timescales for project construction, and the involvement of local organizations in project execution. Importantly, the very projects to be pursued are determined by these consultations. In this way, the community is intricately involved in determining what public benefits will be produced by the QIP (United Nations Reference Harvey, Cd and Fearnley2012, chapter 12). Implemented QIPs are thus projects that civilians want and have chosen for themselves.Footnote 10
Given this, even if a rebel group is motivated to engage in attacks on civilians in an effort to plunder QIP resources, doing so would be problematic for rebel groups that rely upon civilian support for the success of their movement. Victimization in this context would delegitimize the faction as a representative of civilians (Breslawski Reference Breslawski2022). Since QIP provision is negotiated between the PKO and the population, the benefits to civilians are clear, and the violence necessary for rebels to exploit it would be poorly received, given that QIPs map their benefits so well onto the local civilian context. This may be a reason why modestly-sized aid packages determined in consultations with local populations tend to reduce violence (Berman et al. Reference Berman, Felter, Shapiro and Troland2013).
Considering the above, QIPs have important characteristics that disincentivize violence for the purpose of resource extraction to build rebel capacity and incentivize rebels to avoid violence in pursuit of civilian support. QIPs should thus be associated with reduced use of violence by rebels against noncombatant populations. This leads to the following hypothesis:
As QIPs are increasingly invested into districts of civil war states, the number of rebel attacks on civilians in those districts should decline.
Research Design
To test our hypothesis, we require data on the UN’s quick impact projects in ongoing civil wars and rebel attacks on civilians. In the literature linking foreign aid to war violence dynamics, research has moved to subnational levels of analysis. Scholars have argued that both aid and civil war violence are fundamentally micro-level phenomena.Footnote 11 When data was aggregated to national levels, empirical models often yielded inconsistent results. We thus need data on our variables of interest at subnational levels.
Our dependent variable measures attacks on civilians, as coded by the Armed Conflict Location and Event Data (ACLED) Project (Raleigh et al. Reference Raleigh, Linke, Hegre and Karlsen2010). This data identifies ‘violent events where an organized armed group inflicts violence upon unarmed non-combatants’ and includes ‘attempts at inflicting harm (for example, beating, shooting, torture, rape, mutilation, etc.) or forcibly disappearing (for example, kidnapping and disappearances) civilian actors’ (ACLED 2023, 17). While ACLED records instances of violence against civilians by any actor, we only consider violence carried out by rebel groups, which ACLED defines as ‘political organizations with the goal of countering an established national governing regime through violence’ (ACLED 2023, 26). This data is recorded according to the geo-located coordinates of each attack for the day on which the attack occurred, which allows for aggregation to different spatial and temporal units. We measure rebel violence against civilians at time t+1 to address endogeneity concerns.
Our independent variables must capture the intensity of QIP investments in the conflict state. The UN records data on QIPs for several PKOs. Given our focus, the sample is limited to active war. We use data on QIPs from the Multidimensional Integrated Stabilization Mission in the Central African Republic (MINUSCA), the Organization Stabilization Mission in the Democratic Republic of the Congo (MONUSCO), the Multidimensional Integrated Stabilization Mission in Mali (MINUSMA), the Mission in the Republic of South Sudan (UNMISS), and the Interim Security Force for Abyei (UNISFA).Footnote 12 We take this data from the Peace and Security Data Hub (United Nations 2023). The data identify the location, start date, and end date of each project, along with additional information about the projects. Given data availability for our dependent and primary independent variables, our temporal domain is 2017 through 2022.Footnote 13
The fine-grained nature of the ACLED and QIP data allows us to assemble it at subnational levels to the second-order administrative unit in each country.Footnote 14 We also require the data to be structured at a fine temporal level. QIP investments are structured for immediate impact (United Nations Reference Harvey, Cd and Fearnley2012). We thus aggregate QIP data to the month. For our dependent variable, we count the number of independent attacks upon civilians in each district-month.
We note that each QIP investment is commonly seen as a single-shot commitment that will rapidly produce positive community outcomes; QIPs do not include ongoing reinvestments in the same project.Footnote 15 We therefore only record information about QIPs for the months the investment is ongoing. This is when the PKO is actively deploying aid, projects are underway, and combatants and civilians alike can observe the PKO’s engagement in implementing its aid efforts. Information is no longer recorded once the project is handed over to the community.Footnote 16
We generate two variations of our QIP variables to account for the scale or intensity of investment. First, we measure QIPs in terms of their monetary value in US dollars (QIPs-cost), the maximum value of which in a given district-month is $212,799, and the mean value in a district-month is $6,471. This allows us to differentiate across units over time how more or less substantial the UN’s investments have been. As a second measure, we record the number of independent QIPs ongoing in a district month (QIPs-number), the maximum value of which is 32 per district-month and the mean of which is 1.23. Dollar values may not map perfectly onto the overall importance of QIPs to their communities, as the cost may not distinguish between the difference in one costly project relative to multiple simultaneous, though less expensive, investments. We thus count the number of QIPs that are ongoing in each district-month. Both variables are logged, given the positive skew of their distribution. Figure 1 reports the number of QIPs per unit in tandem with the location of attacks on civilians for each country in the data.
Figure 1. Number of QIPs and Attacks against Civilians.
We also include several control variables. First, we address concerns of endogeneity by including the level of rebel violence against civilians at time t and including district-level fixed effects. This allows us to understand whether QIPs in a district are associated with a decrease in violence against civilians, accounting for prior levels of such violence. We include the amount of regime violence against civilians at time t according to ACLED, since violence carried out by the government may shape the rebels’ targeting of civilians (Wood and Kathman Reference Wood and Kathman2015). Ongoing battle events have been shown in previous work to increase victimization, as groups seek to replenish lost resources by preying upon the population following battlefield contests (Hultman Reference Hultman2007; Wood Reference Wood2014). Battles thus records the number of battles that occur between government and rebel forces, as recorded by ACLED. We also control for the duration of PKO presence in each country, as operations deployed for a longer period may be better at protecting civilians and deploy more and more effective QIPs. We control for the number of peacekeepers deployed in each district-month, as the number of personnel may affect PKO effectiveness (Cil et al. Reference Cil, Fjelde, Hultman and Nilsson2020), and larger missions may deploy more QIPs. Finally, we control for the number of actors that used violence in a district-month.Footnote 17 Competition between armed actors is associated with higher levels of violence against civilians (Welsh Reference Welsh2023) and may also deter the implementation of QIPs.Footnote 18
Given the count nature of our dependent variable, we employ a negative binomial model, as linear models often produce inconsistent and biased estimates of counts, and negative binomial models account for heterogeneity and contagion in the data (Long Reference Long1997). We specify our model with district-level fixed effects to account for unit variation not captured by our covariates. We limit our sample to districts that experienced at least one QIP and at least one instance of rebel violence against civilians during the timeframe of the analysis.Footnote 19 Table 1 reports descriptive statistics for each of our variables.
Table 1. Descriptive Statistics
Results and Analysis
We report two models below for each of our two primary independent variables. The first reports the effect for the total dollar value of QIPs invested in a district-month. We report a negative and significant coefficient in Model 1 for QIPs-cost. In other words, as the UN invests increasingly large dollar amounts of aid in its quick impact projects, the number of rebel attacks on civilians decreases.
A similar result is reported in Model 2. Here, we find that as the number of ongoing QIPs increases, civilian targeting by the rebels declines. Together, the results indicate that UN investments reduce the vulnerability of civilians to rebel violence. We theorize that this is due to the unique nature of QIPs, as these projects both disincentivize rebel violence for capacity expansion and incentivize rebels to avoid violence in pursuit of civilian support. The results are summarized in Figure 2, which plots the coefficients of each variable in the model, along with their 95 per cent confidence intervals.Footnote 20
Figure 2. The Effect of QIPs on Rebel Violence against Civilians.
These results are also substantively important. First, we evaluate the results reported on the number of QIPs that are ongoing in a district-month. Holding all control variables at their mean values, we find that to decrease the number of rebel attacks on civilians in a district by 25 per cent, the UN would need to implement two QIPs in a district that currently had no QIPs. We report similar substantive effects when relying on the results for the overall dollar value invested across all QIPs in a district-month. In order to decrease the number of attacks by 25 per cent, the UN would need to spend approximately $60,000 on QIPs in a district that had none.
Additional Analyses
We conduct a number of additional tests to both contextualize and build confidence in the results. First, we address the issue of endogeneity. We begin by examining attributes of the locations to which peacekeeping operations tend to distribute QIPs. We then conduct a rigorous set of analyses to be more confident in the directional relationship that we propose. Second, we run a series of robustness checks to address various additional concerns about our results.
First, we recognize that endogeneity may be a concern for our findings if QIPs are systematically invested in districts in which attacks on civilians are ex ante unlikely. We thus consider analyses that predict which districts to which QIPs are more commonly distributed, using three different independent variables. We use negative binomial models to investigate whether QIPs are more likely to be implemented in district months that are more violent, have more peacekeepers deployed, and have more peacekeeping bases. Predicted values and their 95 percent confidence intervals are reported in Figure 3 below (see Table A2 for regression tables).Footnote 21 Across the three panels in Figure 3, we report that QIPs are more likely to be carried out in district-months that experience more violence, have greater numbers of peacekeepers, and have greater numbers of bases.
Figure 3. District-Month Characteristics of QIPs.Footnote 22
The first panel of Figure 3 offers initial evidence that QIPs are not simply being deployed to districts that are ex ante stable, as more battles predict a larger number of QIPs in given districts. Where battle violence is present, civilian victimization is more common (Hultman Reference Hultman2007; Wood Reference Wood2014). Thus, finding an ameliorative effect of QIPs should be more difficult in our main analyses, given the tendency of QIPs to be invested in relatively dangerous locations. Next, while the second and third panels indicate that QIPs are more likely to be invested in districts where peacekeepers and UN bases are present, the effect of peacekeepers on QIP investment is modest. Further, while bases may be a nominal pretext for investments given the logistical relevance of bases for investment deployment to QIP locations, it is evident that neither bases nor peacekeeper patrols from them are necessary conditions for QIP investments. Rather, as we show in Table 2, many QIPs take place in district months without any peacekeepers present in the district at all.Footnote 23 This is important to rule out the possibility that the relationship we identify between QIPs and violence against civilians is merely due to QIPs being deployed in districts with more peacekeepers, who play a significant role in reducing violence against civilians. Collectively, these results suggest that QIPs are not simply deployed to locations with a lower likelihood of civilian maltreatment at the hands of rebel forces.
Table 2. Percentage of District-Months With(out) QIPs and Peacekeepers (PKs)
We also address endogeneity in additional ways. First, continuing with our exploration of where QIPs are carried out, we regress various temporal lags of differing durations to determine the effect of violence against civilians as a predictor for QIP investment. We use lags of 1, 2, 3, 4, 5, and 6 months between the occurrence of violence and the subsequent deployment of QIPs. None of these lags has a statistically significant negative relationship with QIPs. As such, the UN does not appear to systematically avoid distributing QIP aid to districts with more attacks on civilians. In fact, some of the longer lag times even suggest a positive relationship with QIPs, indicating that QIPs tend to be assigned to district-months that have been experiencing higher levels of violence against civilians (Tables A4–A5).
While these analyses provide a better understanding of where QIPs tend to be assigned, they offer some evidence that district-months in which QIPs are carried out are not comparable to district-months in which QIPs are not carried out. To account for this in the context of addressing endogeneity, we use nearest neighbour propensity score matching, which matches similar ‘treated’ with ‘non-treated’ district-months – or district-months that experienced QIPs and those that did not. This procedure allows us to move closer to inferring the causal effect of QIPs on violence, given the quasi-experimental design afforded by the matching process. We match on all variables used in the main models. After the matching procedure, there were no statistically significant differences between the treatment and control groups across any of the control variables. Using the matched sample, we run three separate models that are identical in specification to those in our main analysis above. Given the balance in observations, we include only our QIP variables: a dichotomous indicator of any QIP investment (our treatment) and our two main variables recording the dollar value of all QIPs invested and the total number of QIPs invested in each district month. We find that our results hold in each model (Table A6).
A number of additional tests also address endogeneity. For instance, we increased the temporal distance between our independent and dependent variables; that is, we measure violence against civilians at time t+2 and t+3 (Table A7) to improve certainty in the temporal order of events. Next, we measure the baseline level of violence against civilians in different ways, including an average from the last three months as well as the number of attacks one year prior. In doing this, we attempt to correct for longer-term spells of violence that might otherwise influence both current violence and QIP implementation (Table A8). In each of these checks, the results for our QIP variables are consistent with those in our main analyses.
Finally, we turn to a series of robustness checks to address various potential concerns. First, we loosen our restrictions on the coding of QIPs, which only record values when QIP implementation is ongoing. To account for the effect of QIPs beyond implementation, we code variables for which the number or cost of QIPs in a given district decays over time with a one-month half-life. Applying a decay function allows us to capture the possibility that QIPs remain consequential following implementation, recognizing that this likely declines with time. Analyses using these variables continue to report largely consistent results (Table A9).
Next, we replace our QIP measure with a measure of the number of all aid projects initiated in a given district month (IATI 2024).Footnote 24 We have made a number of claims about why QIPs in particular are anticipated to reduce attacks on civilians. However, there is a possibility that the presence of peacekeepers causes all types of aid to reduce violence against civilians. Our test reveals that the negative effect on violence against civilians is unique to QIPs, with the relationship between all forms of aid and violence against civilians being null (Table A10). Finally, we include country fixed effects (in addition to district fixed effects) to account for unmeasured differences between countries that may not be addressed by the district fixed effects used in our primary analysis (Table A11). The results do not change meaningfully.
Conclusions
While much of the literature on peacekeeping effectiveness focuses on operational hard power in bringing stability to war-torn countries, UN PKOs can also effectively invest soft power toward peaceful outcomes. The UN engages in quick impact projects in an effort to build support for the operation, its mission, and the peace process. We have argued that these investments are consequential in the UN’s pursuit of its most basic goal: the protection of civilians. While we are limited to the available population of cases on which data are currently available, we find clear evidence that UN aid can help protect the most vulnerable, thereby supporting the more general peacekeeping goal of pursuing stability.
This is an important finding. Had QIPs been found to work at cross purposes to UN protection goals, operations may require significant reform to avoid hamstringing efforts toward peace. No matter how well-intentioned, if UN aid systematically hurts the ability of PKOs to protect civilians, the organization should reconsider their use. Thankfully, we find the opposite, but the policy implications are no less notable. Given that we find that more investments improve civilian protection, an obvious implication is that the UN should invest substantial resources of this sort. Yet, QIPs are often a meagre element of PKO efforts. In fiscal year 2022-2023 alone, MINUSMA’s operating budget in Mali was approximately $1.26 billion. The vast majority of this was dedicated to expenses associated with military personnel, their equipment, and security operations. Only $4.77M was budgeted to cover the cost of ninety-six proposed QIPs (United Nations 2022). These relatively meagre funds can go a long way toward building support for the operation and the peace process. The local peace dividend of protecting civilians in districts to which investments are made is a key factor in pursuing peaceful outcomes. Our results reveal that the UN can do this in meaningful ways with few risks and at relatively low cost.
This finding is especially important given recent peacekeeping participation trends. While research has shown that robustly constituted PKOs can reduce violence and instability, recent work shows that democracies have stepped back from sizeable personnel contributions to PKOs, leaving a void that has been filled by autocratic countries willing to provide blue helmets (Lebovic Reference Lebovic2010; Duursma and Gledhill Reference Duursma and Gledhill2019). This has led some to raise concerns for the future of peacekeeping effectiveness, the stability of ongoing peacekeeping deployments, and the UN’s more general promotion of a liberal democratic order (Levin Reference Levin2021; Melin and Kathman Reference Melin and Kathman2023). An important implication of our research is that even if democratic member-states pull back from providing blue helmets, this does not mean that they cannot meaningfully contribute to core peacekeeping objectives.
Indeed, wealthy democracies bankroll peacekeeping. For instance, for fiscal year 2021-2022, eight of the top ten contributors to the peacekeeping budget were democracies, and they alone provided over 60 per cent of funds (United Nations 2018). One consequential means by which these member-states can remain committed to the liberal democratic order and the protection of civilians is to advocate for expanded budget commitments to peacekeeping and quick impact projects. There is more than one road to peace provision, even when member-states are differentially motivated to engage in peacekeeping efforts.
Finally, future research can build productively upon this work. Whereas the literature on aid and conflict comes to various conclusions regarding the effect of aid on conflict processes, our work suggests that aid can be wielded to protect civilians in certain circumstances. Additional research is needed on this issue to determine how such issues as conflict context, the impartiality or neutrality of the aid provider, or the unilateral or multilateral nature of the provider might more generally affect aid outcomes. Further, one might gain additional traction in assessing the value of QIPs in pursuing a host of additional peacekeeping objectives, including the pursuit of ceasefire agreements, disarmament and demobilization, security sector reform, and democratization, all of which are common objectives of modern peacekeeping efforts. In much the same way that QIPs can be used to secure popular support for the mission toward civilian protection ends, this popular support may also pay peace dividends toward additional peacekeeping objectives.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S0007123425000250
Data Availability Statement
Replication data for this article can be found in Harvard Dataverse at: https://doi.org/10.7910/DVN/8FULWB.
Acknowledgements
We are grateful to Lisa Hultman, who shared the Geo-PKO data through 2022.
Author Contributions
Author names are listed alphabetically. They contributed equally to this research.
Financial Support
None.
Competing Interests
None.
Open access
