Civilian Mindsets and Attitudes toward Peace in Wartime: Evidence from Ukraine

Abstract

Under what conditions do civilians in countries at war support peace settlements? This study develops a theory of civilian attitudes that integrates two major forces shaping wartime thinking to illuminate when people support peaceful compromise. We argue that survival and injustice are two crucial and often competing mindsets that shape how individuals understand and navigate violent conflicts. Civilians exhibiting an injustice-oriented mindset focus more on the objectives of their collective identity group, developing wartime attitudes out of concern for in-group grievances and goals. In contrast, civilians with a survival-based mindset concentrate on the concrete dangers war poses to themselves and their loved ones. To explore these ideas, we fielded two waves of a pre-registered survey in wartime Ukraine in the summer of 2022 and spring of 2023. We find that there is considerable variation across individuals in the extent to which they hold a survival or injustice mindset about the war. Moreover, this variation is strongly linked to their attitudes toward peace across both waves. At the same time, individual mindsets are insensitive to experimental primes, suggesting that they may not be easily manipulated.

Enqué condiciones apoyan los habitantes civiles de los países en guerra los acuerdos de paz? Este estudio desarrolla una teoría en materia de las actitudes civiles que integra dos fuerzas principales, las cuales dan forma al pensamiento en tiempos de guerra con el fin de arrojar luz con respecto a cuándo las personas apoyan el compromiso pacífico. Argumentamos que la supervivencia y la injusticia representan dos mentalidades cruciales, y a menudo contrapuestas, que dan forma a cómo las personas entienden y gestionan los conflictos violentos. Aquellos civiles que muestran una mentalidad más orientada hacia la injusticia tienden a centrarse más en los objetivos de su grupo de identidad colectiva, y desarrollan actitudes bélicas debido a su preocupación con respecto a objetivos del grupo y a los agravios sufridos por este. Por el contrario, aquellos civiles con una mentalidad más centrada en la supervivencia tienden a concentrarse en los peligros concretos que la guerra representa tanto para ellos como para sus seres queridos. Con el fin de estudiar estas ideas, llevamos a cabo dos fases de una encuesta prerregistrada en Ucrania en tiempos de guerra, la primera durante el verano de 2022 y la segunda durante la primavera de 2023. Encontramos que existe una variación considerable entre los individuos con relación al grado en que tienen una mentalidad de supervivencia o de injusticia con respecto a la guerra. Además, esta variación está fuertemente ligada a sus actitudes con relación a la paz en ambas fases. Al mismo tiempo, las mentalidades individuales no son sensibles a las imprimaciones experimentales, lo que sugiere que pueden no ser fácilmente manipulables.

Á quelles conditions les civils de pays en guerre soutiennent-ils un accord? Cette étude développe une théorie des attitudes civiles qui intègre deux forces majeures dans le façonnement de la pensée en temps de guerre pour expliquer pourquoi les peuples soutiennent un compromis pacifique. Nous affirmons que la survie et l’injustice constituent deux états d’esprit cruciaux, qui se font souvent concurrence, quand il s’agit de façonner la compréhension et la gestion de conflits violents au niveau individuel. Les civils présentant un état d’esprit centré sur l’injustice se concentrent davantage sur les objectifs de leur groupe identitaire collectif. Par temps de guerre, ils développent ainsi des comportements spécifiques, car ils se préoccupent de griefs et d’objectifs internes au groupe. Par opposition, les civils dotés d’un état d’esprit de survie se concentrent sur les dangers concrets que la guerre représente pour eux et leurs êtres chers. Pour approfondir ces idées, nous avons mené deux vagues d’enquêtes pré-enregistrées en Ukraine en temps de guerre, lors de l’été 2022 et du printemps 2023. Nous constatons une importante variation d’un individu à l’autre en fonction de la mesure dans laquelle il entretient un état d’esprit de survie ou d’injustice à propos de la guerre. De plus, dans les deux vagues, il existe une forte corrélation entre cette variation et leurs attitudes à l’égard de la paix. Cependant, l’état d’esprit des individus n’est pas sensible à l’amorçage expérimental. Ils ne pourraient donc pas être facilement manipulés.

Introduction

Under what conditions do civilians in countries at war support peace settlements? Broad public support for ending wars is not a guarantor of enduring peace, but it is a critical ingredient. Robust public support for peace helped push the key combatants in Northern Ireland toward a deal in the 1990s (Irwin 1999; Peace Polls N.d.), while declining public support for compromise among Israelis and Palestinians since the failure of the Oslo Process has undermined progress toward a settlement of the Israeli–Palestinian conflict (Palestinian Center for Policy and Survey Research 2020). Moreover, all peace deals are not created equal. Populations may have strong preferences over the actual substance of peace settlements, constraining negotiators’ behavior. Thus understanding the factors behind civilian attitudes toward peace has considerable impact on international politics.

Despite the critical importance of understanding attitudes about peace in conflict settings, we have much to learn about how civilians think and what drives them in war. Some of the most prominent research on civilian attitudes paints them as deeply entrenched in their group identities and approaching new information about the fighting through an intergroup lens (Lyall, Blair, and Imai 2013; Silverman 2019). Yet there are also key models of conflict that stipulate that civilians respond strategically to their treatment in war (Kalyvas 2006; Berman, Shapiro, and Felter 2011), buttressing the intuition that they are driven by a desire to maximize their odds of personal survival–rather than to defend their group identities or redress collective grievances. These ideas appear to be at odds, but there has been little research exploring how they fit together to shape the micro-foundations of conflict in general or mass attitudes toward peace in particular.

We argue that civilians in conflict settings are caught between two competing ways of thinking about conflict and its resolution. One of these, the “injustice mindset,” centers on people’s collective identity groups and their efforts to ameliorate their grievances and achieve social and political justice in the dispute. The other, the “survival mindset,” is focused on the fundamental desire to make sure that one and one’s family successfully navigate the immediate perils of the conflict. The implications of this distinction are simple but powerful: where one falls on the spectrum from an injustice to a survival mindset goes far in accounting for one’s attitudes toward making the sorts of painful concessions required to peacefully resolve conflict.

We test these ideas in two original survey waves conducted in wartime Ukraine.¹ Our survey evidence produces several unique findings. First, our data reveal significant variation in Ukrainians’ mindsets toward the war. Notably, we find that while most Ukrainians (59–68 percent) harbor more of an injustice mindset toward the war, 12–16 percent of the country views the war through more of a personal survival lens. We find important variations in civilians’ wartime mindsets even in a Ukrainian society experiencing a deeply unifying campaign of national resistance against foreign occupation. Second, we find that this variation has considerable purchase in accounting for civilians’ support for peace and willingness to make concessions in both survey waves. And third, we find that Ukrainians were not susceptible to experimental primes designed to alter the intensity of these mindsets. Overall, our results suggest that the distinction between the survival and injustice mindsets is highly relevant to understanding civilian support for peace, even as such mindsets become entrenched and difficult to manipulate in wartime.

Existing Perspectives on Civilians’ Wartime Attitudes

Existing research highlights several factors that shape civilian attitudes during wartime. One critical perspective focuses on people’s attachments to collective identity groups—be they national, ethnic, religious, or ideological—and how social identity shapes wartime attitudes. Indeed, there are many examples of studies highlighting the importance of broad social cleavages and identities in shaping civilians’ perceptions of war. Lyall, Blair, and Imai (2013, 681) show that civilians in Afghanistan tend to perceive abuses by combatants who represent their in-group differently from abuses that are perpetrated by out-groups, providing a “home team discount” to the former but not the latter. Similar studies support these insights by finding that group identification is a key predictor of people’s conflict attitudes, from Kaltenthaler, Silverman, and Dagher (2018) in Iraq and Kaltenthaler and Miller (2015) in Pakistan to Corstange (2019) in Syria and O’Loughlin and Toal (2020) in Ukraine.

A complementary line of research focuses on individual-level variation in civilians’ values, goals, and worldviews in armed conflict. For example, Atran and Axelrod (2008) emphasize individuals’ need to protect their “sacred values”—values that they view as supreme and nonnegotiable—as a driving force in war. Meanwhile, Tellez (2019) shows that the extent to which people have punitive, authoritarian value systems limits their willingness to make concessions to opposing parties. And Shapiro and Fair (2010) highlight how the overarching political aims that civilians prioritize are the strongest predictors of their support for militant organizations in Pakistan. Like the social identity perspective, this research focuses on civilians’ attachments to the different groups in a conflict environment and perceptions of what those groups deserve.

A third line of research centers around the role of conflict experiences in shaping wartime attitudes. Some studies have linked exposure to wartime violence to pro-social attitudes, such as increased cooperation with other civilians and empathy toward in-group members who have suffered (Gilligan, Pasquale, and Samii 2014; Hartman and Morse 2020). Other studies link violence exposure to anti-perpetrator views and a demand for more militarized responses toward out-groups in conflict (Canetti-Nisim et al. 2009; Getmansky and Zeitzoff 2014). One integrative view of these results is that attitudes hinge on the types of violence to which people are exposed (Pechenkina and Gamboa 2020). Other efforts to square these results hold that they represent a desire for risk reduction, given that studies reveal both a greater desire for peace as well as a greater desire for separation from threats among communities at risk (Tellez 2018). This view derives from the argument that people value safety over alternative wartime aims, something which is assumed in influential models of the micro-dynamics of conflict (Kalyvas 2006; Berman, Shapiro, and Felter 2011) but not directly tested.

Finally, research on civilians' behavior in conflict environments—and their attendant survival strategies–also deals substantially with the question of how civilians think in war. This research tends to align with the aforementioned stream on violence exposure in war in centering survival as the prime underlying motivator of civilian thinking (Knuppe 2024). However, research in this area also highlights alternative psychological processes that influence how people think in the throes of conflict, including their responses to violent trauma (Schon 2020) and situational appraisals about their personal circumstances (Milliff 2024).

In sum, past research showcases different factors that shape how people think in war and their attitudes toward peace. However, there has been little attempt to put these paradigms into conversation with one another. Indeed, these perspectives often rest on strong claims about what civilians care about most in conflict. For example, in one study of support for peace in Colombia, Tellez (2018, 1057) claims that “reducing exposure to violence is a first-order priority that takes precedence over identitarian attachments.” In contrast, in an analysis of rumors in Southeast Asian conflicts, Greenhill and Oppenheim (2017, 663) treat worldview as the key force shaping people’s thinking, arguing that “worldview will ultimately be more influential than plausibility” and “social truth will trump objective truth” in how people process new information in war. Given the strong reasons to think that both identities and values as well as security and violence are powerful influences on civilian motivations, there is a need to understand the extent to which civilians perceive war through alternative points of view.

A Theory of Civilian Mindsets during Wartime

During wartime, we argue that civilians are caught between two key modes of thinking (or “mindsets”) which shape their specific attitudes, beliefs, and behaviors. On the one hand, individuals engage in collective injustice-based thinking about war. Drawing on the salience of social identity, this mindset focuses on the group-level aims that individuals want to pursue. These goals may be nationalist in nature, such as Ukrainians wanting to defeat Russia’s occupation, or they may correspond with sub-national identities, such as Sunni Arabs in Iraq wanting meaningful representation in the government during the onset of the Islamic State occupation in June 2014 (Silverman, Kaltenthaler, and Dagher 2021).

Injustice-based thinking centers around individuals’ moral and ideational goals during wartime. It deals with what people think is fair and right in terms of the war’s outcome. It also builds on the longstanding “grievances” tradition in the study of civil war, which emphasizes communal groups and the injustices that their members want to see redressed in order to resolve conflict (Gurr 1993). While some macro-level conflict studies cast doubt on the importance of grievances (Fearon and Laitin 2003; Collier and Hoeffler 2004), recent work suggests that they play a key role in explaining where conflicts break out and how they unfold (Cederman, Gleditsch, and Buhaug 2013; Masterson 2022). Related research that is closely tied to this injustice mindset includes studies highlighting the role of sacred values and their violation in risky political behaviors (O’Brien 1996; Jasper 1998; Dornschneider 2011; Ginges et al. 2011).

On the other hand, individuals also engage in survival-based thinking. Survival-based thinking focuses on people’s basic desires to see themselves and their loved ones safe from harm. War is full of peril, and those in its crosshairs have a strong desire to make sure they and their families will make it out alive and unharmed. Indeed, the basic need for physical safety is seen as a “first-order priority” for civilian populations by some analysts (Tellez 2018, 1057–58), particularly in the literature on the effects of exposure to wartime violence. Such arguments harken back to classic models from social psychology in which fundamental needs like survival were seen as the strongest engines of human action when left threatened or unfulfilled (Maslow 1943; Lester 2013), and builds on more recent experimental work showing that people’s social and political preferences can be overridden by sufficiently large material incentives (Brandimonte and Ferrante 2015; Bursztyn et al. 2020). Likewise, rationalist theories of war analyzing how combatants’ treatment of civilians can win (or lose) public support, model civilians as primarily passive, survival-driven agents (Kalyvas 2006; Berman, Shapiro, and Felter 2011).

We argue that these two mindsets function as competing tentpoles that pull people in different directions when they reflect on the prospects of a negotiated settlement. To achieve a settlement, individuals must abandon their quest for total victory and enter negotiations with the adversary. Peace requires painful compromises on ostensibly righteous goals and concessions to adversaries whose goals are perceived to be inferior. Thus, individuals who indulge in injustice-based thinking are likely to resist compromise when they identify with one or more groups and share or support their goals in the conflict, be those liberating an occupied territory, toppling an unjust regime, or stamping out a security threat.

Examples of injustice-based thinking in wartime abound. For example, many Armenians and Azerbaijanis support their countries’ aims of exercising control over the contested area of Nagorno-Karabakh, seeing their claims to the land as just and abandoning them for peace as unacceptable (Melander 2001). Likewise, in the case of Ukraine, injustice-based thinking has by all accounts played a powerful role in shaping attitudes toward the country’s fight with Russia, with Ukrainian civilians often articulating ideas like “I do not want my children to live in Soviet Union 2.0, [but] in a free democratic country”² and “We need to oppose Russia...We have chosen, not a Russian path, but a European one”³ as crucial frames for understanding their struggle. Or, as one Ukrainian colorfully described their reaction to Russia’s pre-2022 military incursions in Ukraine, “If someone took over your kitchen and started frying cutlets there...what would you do, pat them on the head?...We need to support our homeland.”⁴ Such sentiments are reminiscent of a Palestinian analogy—that of finding one’s house invaded and then being asked to split it with the invaders—that some Palestinians voice as opposition to a two-state solution (Tessler 2019).

In contrast, individuals drawing on a survival mindset will be more likely to support political compromises and peace settlements. War is deadly for civilians, threatening their lives and the lives of their loved ones if they are either victimized by one side or caught in the crossfire. Individuals who are most concerned with survival are more likely to welcome peace, which promises to reduce the risk of being killed and support this fundamental survival imperative for them and their families.

Evidence for survival-based thinking during wartime is common. For example, a Pakistani journalist notes that for civilians in the tribal areas where US drone strikes frequently occur, “drones are not just some abstract talking point. Just getting through the day has become a high-stakes game” (Mustafa 2013). Likewise, despite the powerful draw of injustice-based thinking for many in the war in Ukraine, some Ukrainians—particularly those in the hardest hit areas—appear to be animated above all by concerns for their immediate physical safety. For instance, as one civilian who fled Mariupol—a city largely destroyed during Russia’s siege at the start of the war—explained to a reporter: “I focused on just survival. We didn’t know if it would be possible to ever leave. I did not believe escape would be possible.”⁵

However, a few caveats about the interaction of survival and injustice-based mindsets are worth highlighting. First, these mindsets do not operate as binary switches, but as lenses that people can adjust when looking at a conflict with varying degrees of intensity. Second, wartime mindsets are not mutually exclusive—individuals can and do engage in both types of thinking as they face violence threats. For example, someone might be focused on survival and still be far more supportive of one side’s war aims. Yet, we argue that there are differences in the rank ordering of an individual’s preferences at a specific point in the conflict. Third, we recognize that both outlooks can fluctuate over the course of the conflict and with people’s idiosyncratic circumstances. We simply posit that they contain some signal along with any noise that may exist and that people’s wartime attitudes of all kinds fluctuate, yet that has not precluded scholars from viewing them as key objects of study.

While our framework breaks new ground in the study of civilian wartime attitudes, it also builds on research in other security domains. For example, research in criminology has explored the extent to which people’s judgments about the appropriate punishments for criminal behavior derive from moral considerations, such as deservedness and moral culpability (Carlsmith, Darley, and Robinson 2002). Meanwhile, work on attitudes toward corruption in insecure environments examines the degree to which individuals who face security threats from criminals will tolerate government corruption within their communities if they think it can “keep the peace” (García-Ponce, Zeitzoff, and Wantchekon 2021, 709–10). Finally, studies of transitional justice also grapple with these dilemmas, looking at how citizens navigate the tradeoffs between their desire to punish wartime abuses and achieve justice versus the risk that such preferences could threaten their own hard-won security (Samii 2006; Bratton 2011; Daly 2022). We bring these tradeoffs and their consequences into the domain of protracted wars—where we argue that they can help integrate disparate ideas about how civilians think about war and the prospects of peace.

One might worry that our framework is too causally proximate to the outcomes of interest to be useful, and should instead be understood as something more tautological in nature. Yet it is not a priori clear that civilians who are concerned chiefly with survival should support peace, whether in Ukraine or beyond. Indeed, depending on whether people perceive a potential peace deal as an institution that is reliable enough to deliver security, or as a dangerous form of appeasement that could embolden their adversary, prioritizing survival might push them in distinct directions.⁶ This question is well-grounded in the broader conflict studies literature, including in scholarship on the challenges of credibly committing to end conflicts (Walter 2002), the difficulty of designing effective power-sharing agreements (Hartzell and Hoddie 2003), and the idea of ethnic conflicts hamstrung by security dilemmas (Posen 1993).

A potential Russo–Ukrainian peace agreement is no easy case in this regard. Many Ukrainians have a profound mistrust of Putin’s intentions given his record of irredentism and military aggression toward neighboring states (Slantchev and Goemans 2024). In particular, Ukrainians close to the conflict’s frontlines might have a particular reason to care about their survival, but be wary of a peace deal (since their communities could be attacked or even seized the most easily if the opponent reneged). In sum, the question of whether survival-driven civilians are more likely to support peace is not ex ante obvious, either in Ukraine or elsewhere. Rather, results showing that survival-driven individuals are more supportive of peaceful settlement would suggest not only that such people are willing to compromise, but also that they believe that a non-trivial settlement is actually possible.

We test the ideas in our theoretical framework by observationally measuring civilians’ survival and injustice mindsets and using them to explain attitudes toward peace, as well as experimentally priming them with relevant messaging. The foregoing theoretical discussion combines with our approach to produce the following, pre-registered hypotheses:  

 

Building on literature on how direct exposure to violence alters preferences toward peace (Tellez 2018; Pechenkina and Gamboa 2020; Pechenkina and Argo 2020), we also pre-registered the hypothesis that exposure to violence will predict greater support for compromise. We viewed the exposure-compromise linkage as bound up with the degree to which people hold a survival-oriented mindset. This logic suggests the following hypothesis:  

The Case of Wartime Ukraine

We investigate these hypotheses by conducting a two-wave survey in Ukraine during its ongoing war against Russia. Ukraine is useful to study because it represents a hard case for our argument. Ukrainian society has experienced a powerful rally effect around the country’s threatened sovereignty and identity in the face of Russian aggression. Indeed, polls have shown overwhelming support for the war among Ukrainians, robust approval of the country’s top leadership, and limited appetite for territorial concessions (International Republican Institute 2022; Gallup 2022), especially during the period under study (the first year of the war). Academic research shows that Ukrainians tend to view the struggle for national independence as a “categorical imperative” rather than a pragmatic bargain (Dill et al. 2023, 2–3). Ukraine’s early success against Russian forces likely facilitated this rally effect as citizens were able to see not just the righteousness but the viability of national resistance. In this context, it should be hard to identify substantial variation in civilian mindsets. Rather, one would expect an injustice mindset to predominate in wartime Ukraine in 2022–23, with a population united behind a relatively successful war of national resistance.

Ukraine gained its independence from the Soviet Union in 1991, and in the ensuing years, the country grew deeply polarized about whether it would benefit from greater integration with the EU or Russia. The east-west orientation of foreign policy represented the main cleavage in Ukrainian politics, on which all political parties could be positioned. In November 2013, Ukraine’s then pro-Russian president Yanukovich announced economic integration with Russia as opposed to the EU, which sparked peaceful protests. After a brutal police response, the protests turned violent, and the pro-Russian president fled for Russia.

Between February and March 2014, Russia annexed Crimea, exploiting the turmoil that had broken out in Kyiv. Later that spring, peaceful anti-interim government protests occurred in most major cities of eastern Ukraine. Some of those protests, including those in Donetsk, Luhansk, and Slovyansk, turned into forced seizures of administrative buildings, police headquarters, and military bases by what would become anti-government militia members and/or Russian volunteers (Rachkevych 2014). The militias included local residents, whose goals ranged from independence to an autonomous region within Ukraine, and whose only unifying aim was to hold a electoral referendum (Chivers and Sneider 2014). These events ignited the Donbas War, which pitted Russian-backed rebels and volunteers against Ukrainian state forces. The war nominally ended in January 2015 with the Minsk peace agreement, but there were repeated ceasefire violations and fighting continued after it was signed. From 2014 to 2021, Russian-sponsored conflict in the Donbas killed “at least 3,404 civilians, 4,400 Ukrainian forces, and 6,500 members of other armed groups,” with the bulk of these deaths coming between 2014 and 2015 (Office of the UN High Commissioner for Human Rights 2022, 3).

The annexation of Crimea and instigation of the Donbas War solidified Ukrainian identity, easing the country’s longstanding east–west political divide. In two post-Maidan presidential elections, Petro Poroshenko (who received 55 percent of the vote in May 2014) and Volodymyr Zelensky (who received 73 percent of the vote in March 2019) won the majority of Ukraine’s provinces, with no sharp disparity between northwest and southeast. After a brief honeymoon period, tensions with Russia spiked under Zelensky. This occurred for several reasons, some of which include Zelensky resisting Putin’s push to implement the Minsk agreement (as Putin saw it), Ukraine’s 2019 state language law rolling back the use of Russian language in Ukraine’s schools and media in 2020–2021, and the shutdown of oligarch Viktor Medvedchuk’s pro-Russia television channels in Ukraine (Medvedchuk being a close friend of Vladimir Putin). These events coincided with Putin’s declining popularity, which often incentivizes authoritarian leaders to create or exploit external threats in order to shore up their rule. In this context, 2021 saw a dramatic escalation in Russia’s rhetoric toward Ukraine and the launching of extensive military exercises on Ukraine’s border. Russia ultimately initiated a full-scale invasion of Ukraine on February 24, 2022, launching the bloodiest war in Europe since World War II.

Data and Research Design

Survey Description

We test our hypotheses on two original survey waves fielded in wartime Ukraine. The Kyiv International Institute for Sociology (KIIS)—the leading independent public opinion firm in the country—fielded the questionnaires by cellular telephone based on a simple random sample of Ukrainian citizens across the country.⁷

KIIS uses a computer-assisted telephone interview approach to conduct surveys. No pilot data were collected for this study. KIIS enumerated Wave 1 in July 2022 as a phone interview with a random sample of 2,000 respondents. Wave 2 was enumerated in February 2023 among a second, representative sample of 2,007 respondents. All Ukrainians over 18 years old within the country were eligible for recruitment. KIIS enumerators described the research as a study on public opinion and obtained verbal confirmation of informed consent before the start of the questionnaire. KIIS researchers and the authors verified survey questionnaire translation between English, Russian, and Ukrainian.

The survey design proceeded as follows. Using software, the KIIS generated random cell phone numbers. After removing non-existing phone numbers, 32,470 and 31,979 phone numbers were randomly selected and contacted until 2,000 and 2,007 complete responses were obtained in waves 1 and 2 respectively. The contact rate was 56 percent and 49 percent, the cooperation rate was 11.5 percent and 13.6 percent, and the response rate was 6.8 percent and 7.0 percent in the two respective waves.⁸

The research design was intentional in ensuring the ethical enumeration of both survey waves. For example, to avoid possible safety concerns, we abstained from surveying areas under Russian control (e.g., Luhansk, Donetsk, and Crimea). Participants were asked to provide their consent before the survey and were told they were free to withdraw at any time while it was underway. Participants’ cell phone numbers were not saved and linked to their answers in order to protect their anonymity, maximizing security and confidentiality in the study.⁹

Waves 1 and 2 include 2,000 and 2,007 randomly selected respondents from 709 and 754 unique communities (hromada) across Ukraine’s 22 and 23 oblasts respectively. The residents of areas under Russian control in July 2022 and February 2023 such as Crimea, southern Kherson, southern Zaporizhzhya, southern Donetsk, and Luhansk oblasts were excluded for security reasons. The sample of wave 1 (wave 2) was 46 (44.1) percent male, with a mean age of 47.8 (49.3) and s.d. = 16.1 (16.4). 94 (95) percent of respondents identified as ethnic Ukrainians, though only 72 (78) percent recorded that Ukrainian was the primary language spoken in their homes. Seventy-one (80) percent of the sample completed some college or vocational training. Two-thirds (three-quarters) of respondents were employed and nearly 90 (95) percent reported a monthly income that allowed them to meet daily basic needs. Since the conflict began on February 24, 2022, 14 (15) percent of respondents registered being internally displaced by conflict-related violence, while 86 (85) percent currently reside in the same community as before February 2022. The Online Appendix includes a complete list of descriptive statistics.

Figure 1 visualizes the spatial distribution of respondents across survey waves, illustrating that the sample achieved wide territorial coverage across Ukraine. The shading indicates a cumulative count of violent events per district prior to each wave, while the counts indicate how many interviews were enumerated in said district.

Dependent Variables

Three different survey questions measure respondents’ support for a settlement of the war with Russia, resulting in five distinct outcomes. Table 1 summarizes the primary dependent variables across each survey wave.

Willingness to Save Lives over Keeping Territory

First, Save_Lives is a four-point ordinal measure, in which higher values indicate greater general willingness to save lives rather than keep territory based on the following question: “Do you think it is more important to reach a peace deal soon to save Ukrainian lives even if it means Russian control over some Ukrainian territories, or to fight a longer war to return the territories that Russia occupied during its 2022 invasion?” Four answer options ranged from “Much more important to save lives than to keep territory” to “Much more important to keep territory than to save lives.”

Willingness to Compromise with Russia

Second, Compromise_Binary asks respondents about whether they would support surrendering Ukrainian territory in order to end the conflict: “To achieve peace and preserve independence as soon as possible, Ukraine can give up some of its territories.” Unlike Save_Lives, Compromise_Binary, offers respondents a binary choice between “agree” and “disagree.”

Finally, we asked respondents whether they would support a series of concessions to Russia to end the conflict including: “Which of the following ideas do you consider acceptable to achieve peace with Russia as soon as possible and save Ukrainian lives as part of a peace deal with Russia? Please answer yes to all that apply.” Respondents gave yes or no answers to the options after this question. The order of concessions was randomized to avoid question-order effects.

Figure 2 summarizes responses across both waves.¹⁰ The inter-item correlation among the four questions related to territorial compromise is high (0.67). As one would intuitively expect, the more extensive the territorial concession to Russia, the lower its support. For example, more individuals (12 percent) are willing to accept Russia’s control over Crimea than over Kherson (2.7 percent). Using inverse covariance weighting, these four items were combined into an index Territorial_Compromise (Blair, Grossman, and Weinstein 2022).

We then use the NATO and Russian language items as standalone outcomes, since they do not have a high correlation with territorial items or with each other. Forty-three and 33 percent of respondents in waves 1 and 2, respectively, reported being willing to compromise on Ukraine’s neutral status and committing to no future participation in the NATO alliance (the resultant binary outcome is No_NATO). Twenty-four and 16 percent of respondents in waves 1 and 2 respectively said they would be willing to accept the return of the Russian language in secondary schools as the primary language of instruction at pre-2020 levels as a concession to end the fighting (the resultant binary outcome is Language_Policy).

Independent Variables

We include three independent variables testing H1 and H3 in our main empirical model since we used these observational measures in both waves. In contrast, we tested H2 with a survey experiment in the first wave only, so we analyzed it only in the results section. Descriptive statistics for our main independent variables are listed in Table 2.

Natural Variation in Civilians’ Wartime Mindsets

To test hypothesis 1, we observationally measured respondents’ injustice versus survival mindsets using the following question:

Imagine that you can allocate 100 total points between the following two answers. Which of the following is more important to you when you think about the war with Russia? For example, you can allocate 85 points to answer 1 and 15 points to answer 2, or 15 points to answer 1 and 85 to answer 2.

Option 1: The war is important because it determines whether my family and I live or perish.

Option 2: The war is important because it determines whether Ukraine can free itself from Russian oppression and domination.

We constructed the variable Survival_Mindset which ranges from 0 to 100 points and captures the extent to which respondents prioritized concerns about perishing from warfare over Ukraine’s freedom from Russia (0 = only concerned about Ukraine’s freedom, while 100 = only thinking about personal survival). Survival_Mindset captures respondents’ natural inclination to see the war through the lens of personal survival versus injustices against their in-group.

In wave 1, 68 percent of the sample held the injustice mindset and only 12 percent held the survival mindset; 19 percent were indifferent between the two. The second wave in February 2023 yielded very similar averages (59 percent held the injustice mindset, while 16 percent held the survival mindset). Spatial variation in the subsample remained in the same direction as in our survey in July, implying that these mindsets do not shift rapidly—at least in the absence of dramatic changes on the battlefield (wartime violence has been geographically clustered in Ukraine’s southeast since April 2022).

Experimental Priming to Induce Injustice or Survival Mindsets

To evaluate hypothesis 2—whether we can prime survival and injustice mindsets through specific messaging—we divided the respondents of wave 1 into three groups. The first group was randomly assigned to the control group and received no prime. The second group was assigned to a treatment designed to prime one’s survival mindset with the following language:

As you know, there are many reasons why this war is important. One reason is because it can directly impact your survival–whether you or your family live or die. Over the last four months, Ukrainians have had to make many decisions about what to do in order to survive, including where they can safely go in their neighborhood, whether to flee or stay in their homes, and whether to join or assist the Territorial Defense Forces. In the past four months, how often have you viewed the war through this lens of your and your family’s survival?

Respondents answered the question on a four-point Likert scale from “not once” to “constantly.” We use a binary indicator of receiving this treatment (Survival Treatment: Binary) in our analysis of the experiment’s effects. The third group was assigned to a treatment designed to prime an injustice mindset:

As you know, there are many reasons why this war is important. One reason is because it determines the fate of Ukraine as it attempts to free itself from Russian oppression and domination. Over the past century, Ukrainians have experienced many injustices against their country from Russia, including the Holodomor, the persecution of prominent Ukrainian freedom fighters and intellectuals in the Soviet period, and Russian attempts to subjugate Ukraine and meddle in its elections after independence. In the past four months, how often have you viewed the war through this lens of Russia’s injustices and many crimes and against Ukraine?

Wave 1 respondents answered the question on a Likert scale ranging from “not once” to “constantly.” We use a binary indicator of receiving this treatment (Injustice Treatment: Binary) as well.

Exposure to Wartime Violence

To test hypothesis 3, we include three measures of exposure to wartime violence. Objective measures rely on respondents’ self-reported location, while subjective ones rely on their perceptions about their exposure to wartime violence and its impact on those around them.

First, we use the binary indicator Frontline_Oblast to capture whether respondents’ self-reported location was within a province that saw major battles in the month prior to our survey (Donetsk, Kharkiv, Mykolaiv, and Zaporizhia). Luhansk and Kherson oblasts were excluded from wave 1 because they were almost entirely controlled by Russia and no respondents were recruited from those provinces. In wave 2, however, northern Kherson was under Ukraine’s control and so we included it. For robustness, we recoded Frontline_Oblast as 1 for Dnipropetrovsk and Odesa oblasts—results are stronger under such coding.

Second, we include event data collected by the Violent Incident Information from News Articles (VIINA) project (Zhukov 2022). VIINA records district-level shelling, bombing, and other violent events initiated by the Russian side during the two months before each survey wave.¹¹ The Online Appendix displays the distribution of conflict events by type and survey wave, reflecting that Ukrainian civilians have been most frequently exposed to artillery shelling.

Third, we include two self-reported measures of exposure to wartime violence. One question focuses on personal experience in the war and the other addresses the experience of one’s family and close friends. Each respondent answered “yes” or “no” to the six forms of wartime violence (order was randomized). The Online Appendix visualizes the resulting variables; those starting with “personal” reflect respondents’ first-hand exposure to conflict, while those beginning with “family” capture civilians’ indirect experience based on what happened to their family and close friends in the war. A median respondent reports that they personally experienced one event and indirectly experienced two events.

Control Variables

To avoid post-treatment bias, we only include those controls that could not be affected by exposure to warfare (Montgomery, Nyhan, and Torres 2018). As such, all regression models include respondents’ gender, age, education, employment, income, displacement status, and ethnicity.

Model Specification and Estimation Strategy

We employ multi-level models to analyze the relationship between survival mindset and support for wartime settlements because a multi-stage sampling design drawing on a national census shaped the enumeration of each survey wave. Specifically, we estimate a Multi-Level Model for Change (MLMC) by including random effects for oblast and survey wave (Gelman and Hill 2007; Cernat 2023). The upshot is a model that allows us to examine variation across key variables for respondents within and across oblasts in each survey wave. As such, our analysis is based on various specifications of the following model:

where:

• |$y_{it}$| represents the outcome of interest—either Save_Lives|$_{it}$| or Compromise_Binary|$_{it}$| – for respondent (⁠|$i$|⁠) at survey wave (⁠|$t$|⁠).

• |$\gamma _{00}$| includes a fixed effect for the intercept, representing an outcome’s grand mean over all respondents in wave 1.

• |$\gamma _{10}$| represents an outcome’s rate of change between waves 1 and 2. Wave is a binary variable.

• |$\xi _{0i}$| includes a random effect modeling variation between oblasts in wave 1.

• |$\xi _{1i}$| includes a random effect for time, indication variation in the rate of change between waves 1 and 2.

• |$\xi _{2i}$| is a random effect modeling Survival_Mindset across oblasts. By definition, the model includes an estimate of the fixed effect of Save_Lives.

• |$\epsilon _{it}$| is the residual, indicating how much respondents vary around their average value of each outcome.

We include fixed effects for time-invariant variables, such as respondents’ demographic characteristics. The model with Save_Lives includes a four-point ordinal dependent variable; models with Compromise_Binary, No NATO, and Language_Policy use binary dependent variables; the model with Territorial_Compromise includes a continuous outcome.¹²

Empirical Results

Table 3 presents the main results for each outcome of interest.¹³ Our five dependent variables are displayed in the table’s five columns, and labeled accordingly. Our primary independent variables include measures of exposure to wartime violence: Frontline_Oblast, Violent_Events, Personal_Victimization, Family_Victimization (designed to test hypothesis 3) and Survival_Mindset (designed to test hypothesis 1). Fixed effects for respondent-level covariates are included in the supplemental analysis in the Online Appendix.

Natural Variation in Civilian Mindsets

We begin our analysis by examining how survival and injustice mindsets affect civilian attitudes on saving lives and compromising. Hypothesis 1 predicts that respondents with a survival-oriented mindset will be more willing to settle the conflict than those possessing an injustice-oriented mindset.

We first examine the distribution of survival mindsets across each survey wave (these distributions are visualized in the Online Appendix). Roughly 90 percent of respondents across waves answered the mindset question. The average respondent in both waves viewed the conflict as substantially more about the protection of Ukraine’s sovereignty (scores lower than 50) than the protection of their personal survival (scores greater than 50). In wave 2—fielded in February 2023—the distribution is less right-skewed (that is, skewed toward the survival end of the spectrum).

Table 3 shows that there is a strong pattern consistent with our expectation in H1: naturally occurring variation in survival and injustice-oriented mindsets tracks well with support for peace settlements across respondents. All five multilevel models for change estimate the association between survival mindset and wartime settlement, controlling for survey wave. We find that stronger embrace of personal survival (versus collective injustice) is associated with significantly greater support for saving lives over keeping territory, including making territorial concessions, eschewing membership in NATO, and allowing more Russian language education in Ukrainian schools. Consistent with H1, this suggests that the injustice and survival mindsets have meaningful purchase on wartime attitudes.

To convey the substantive impact of this finding, Figure 3 shows the predicted probabilities of two key outcomes—Save_Lives and Compromise_Binary)—by the degree to which respondents hold a survival mindset about the conflict. These relationships appear among both frontline residents as well as those living far from combat. As is apparent, support for conceding important contested territories in the war varies strongly with the degree to which an individual perceives the conflict through a survival as opposed to an injustice lens.

Panel (a) of Figure 3 shows the predicted probabilities of choosing each of the four answer options of the ordinal variable Save_Lives. Consider the bottom right plot, displaying the probability of answering “much more important to save lives over territory.” A hypothetical respondent who is halfway between being indifferent and fully embracing the survival outlook (i.e., Survival_Mindset = 0.75) is predicted to value saving lives 47 percent of the time, which is 20 percentage points higher than a frontline resident who is halfway between being indifferent and fully embracing the injustice mindset (i.e., Survival_Mindset = 0.25); this difference is 18 percentage points for the same hypothetical individual residing off the frontline. A hypothetical individual living in the frontline oblast, who is “all in” on the survival mindset (Save_Lives = 1.0) is predicted to favor saving lives 57 percent of the time, which is 37 percentage points higher than a hypothetical solely injustice-focused frontline resident. Finally, comparing an individual outside the frontline who only focuses on injustices against Ukraine (Survival_Mindset= 0.0) to someone outside the frontline who views the war exclusively through the survival lens (Survival_Mindset = 1.0) yields a jump in one’s willingness to save lives of 34 percentage points. These are substantively large changes.

Panel (b) of Figure 3 displays the predicted probabilities of respondents agreeing with the statement “To achieve peace and preserve independence as soon as possible, Ukraine can give up some of its territories” (i.e., Compromise_Binary). Similarly, large differences are apparent. For example, a hypothetical individual living in the frontline oblast, who is “all in” on the survival mindset (Save_Lives = 1.0) is predicted to favor territorial compromise 56 percent of the time, which is 35 percentage points higher than a hypothetical solely injustice-focused frontline resident. Meanwhile, comparing non-frontline “all-in” survival-versus-injustice focused individuals yields a difference of 29 percentage points. These results highlight the impact of the survival versus injustice mindset distinction in substantive terms.

Exposure to Wartime Violence

We next turn to the effects of exposure to wartime violence on public attitudes toward compromise to end the war (H3). Before interpreting these results, it is worth noting the descriptive differences among observational and subjective measures of violence exposure. The Online Appendix displays a heat map of all the correlations among the violence-exposure variables. As one would expect, there is a strong correlation among the indices measuring personal and family exposure to wartime violence (0.73). We also observe a modest correlation (0.34) between frontline oblasts and the cumulative level of district-level violence. Interestingly, there is only a weak relationship between the subjective measures of violence exposure and district-level violent events (0.23 and 0.25). Weak correlations suggest that personal exposure to violence varies widely among civilians living in particularly violent districts or that respondents may misreport their experiences.

Table 3 shows that subjective indicators of exposure to violence have no meaningful effect on one’s willingness to save lives or to compromise (inconsistent with H3). By contrast, the Frontline_Oblast indicator has a strong positive association with one’s support for saving lives or compromising on various issues (p-value |$<0.054$| in Model 3). Russian-Initiated_Combat_Events, however, generally do not have an association with outcomes–with the sole exception of Model 5, which indicates that higher frequency of combat events increases one’s willingness to compromise on the issue of allowing secondary schooling in the Russian language.

Overall, we find that subjective measures of violence exposure have little impact on outcomes, while objective measures yield mixed results. There is a positive association between residing in a frontline oblast and willingness to save lives and compromise (as expected by H3), but almost no association exists between cumulative combat events and outcomes (contradictory to H3). Ultimately, the results do not provide strong and consistent evidence for war exposure-based theories of support for peace, although they suggest that spatial proximity to the frontlines may have a stronger connection to peace attitudes than either more general event-based measures or subjective measures of violence exposure.

Experimental Priming of Civilian Mindsets

We then move to test of hypothesis 2: that priming the respondents to view the war through a survival or injustice lens will affect their views on peaceful compromise with Russia.

In contrast to the clear observational connection between civilian mindsets and peace attitudes, the Online Appendix shows that experimentally priming survival or injustice-oriented mindsets had no systematic impact on any of the five outcomes of interest. We thus reject H2, finding no evidence that experimental manipulations of the two mindsets shape attitudes toward saving lives or compromising with Russia. While it is difficult to conclusively interpret this null result, it points toward the ways in which these deep mindsets may become firmly entrenched and difficult to manipulate in wartime. We engage with the question of why the primes did not change respondents’ attitudes further in the conclusion.

Alternative Explanation: Language and Ethnic Identity

Our primary result so far is that civilians who exhibit the survival mindset are more likely to support a painful compromise with the enemy (H1). Yet, a skeptic might respond that our main findings are primarily driven by social identity, especially among those with a strong attachment to the Russian language or ethnic identity. In other words, it might be the case that Russian-speaking Ukrainians would be more likely to hold a survival mindset about the war to begin with and more likely to support compromises to end it due to differences in their worldviews or conflict experiences. Does the effect of survival mindset on compromise persist if we hold respondent identity constant? That is, do we observe similar differences in support for compromise across those with different mindsets when we compare Russian and non-Russian identifying respondents?

Table 4 reveals that differences-in-means in support for saving lives and territorial compromise between Russian speakers with an injustice and a survival mindset are substantively and statistically significant. Russian speakers who embrace the injustice mindset are on average 0.78 standard deviations less likely to choose saving lives over keeping territory, and 0.83 standard deviations less likely to choose painful compromise. These differences persist if we compare different mindsets among those who identified as Russian speakers, and when we add those who said they speak both Russian and Ukrainian or found it hard to say what language they spoke at home. Finally, comparing ethnic Russians with different mindsets generates large substantive differences in willingness to save lives or concede territory. However, the results are not statistically discernible for territorial compromise in wave 1 and for both outcomes in wave 2 because of the low power of the test.

Discussion and Conclusion

On December 22, 2023, the Palestinian photojournalist Motaz Azaiza—who has gained considerable online influence by documenting human suffering in Gaza amid the Israel-Hamas war that started in October of that year—posted a video of a man in Gaza’s Bureij refugee camp giving an impassioned speech about his views on the conflict.¹⁴ “We are a people who want to live!” shouts the man in Arabic during an extended monologue about how he and other Palestinians have had enough of death and suffering over the past 75 years. “We want to live the good life like the Israeli people. We are tired as the Israelis are as well.” This video quickly went viral, reaching five million views and generating sharply divergent reactions among Palestinians and Arabs. Some lauded the man’s earnest desire for Palestinians to live in safety and freedom, while many others attacked him and Motaz for posting content that could be interpreted as undermining Palestinians’ appearance of resolve in the fight against the Israel Defense Forces.

Motaz’s post—and the reactions it garnered—exemplify the arguments advanced in this study: civilians in war vary in the extent to which they hold more of a survival or an injustice mindset about conflict, and that this variation explains the degree to which they will consider backing painful compromises in the name of peace. The Gazan man filmed by Azaiza and those who promoted his views clearly looked at the conflict through a survival-oriented lens, while those who rushed to denounce the video were doing so through an injustice-oriented mentality.

We tested these ideas with a pair of surveys fielded in wartime Ukraine in July 2022 and February 2023. These surveys yielded four key results. First, we find significant naturally occurring variation in the extent to which civilians held survival versus injustice mindsets. While most Ukrainians harbored more of an injustice mindset, a nontrivial portion viewed the conflict through a survival lens—notable given its immensely popular cause and considerable degree of success (during the period of interest) from the Ukrainian perspective. Second, observed variation was strongly linked to civilians’ support for peace and willingness to compromise with Russia across different outcomes. Third, respondents’ attitudes toward peace were not sensitive to experimental manipulations designed to alter those mindsets. Finally, residing in a frontline oblast (but not other measures of exposure to violence) predicted greater support for compromise.

How do we parse these findings? Our results strongly suggest that civilians’ general mindsets or outlooks on a conflict—that is, whether they understand it through more of a survival or an injustice lens—are powerful forces that undergird their wartime attitudes. In particular, we find a strong and robust association between thinking of the war in terms of survival and support for peaceful compromise to end the fighting. This finding is significant in part because caring about survival could have pushed civilians in other directions, such as making them more pro-war if they saw the conflict as an inescapable security dilemma in which peace agreements would be exploited by the opponent to threaten them further (Posen 1993; Luttwak 1999). Instead, our results show that a survival mindset makes civilians more supportive of a peaceful settlement to the conflict, suggesting that they (1) are willing to trade off collective political ambitions for personal and familial security and (2) do not reject the idea of peaceful compromise with the adversary in terms of its ability to actually stick and deliver that security.¹⁵

Our findings also raise questions about the origin of both mindsets. While beyond the scope of this study, our analyses suggest a key insight about the deeper contours of these mindsets: that they appear to be very “sticky” and hard to move or manipulate after sustained exposure to war. Two facts lead to this conclusion. First, our efforts to prime the two mindsets completely failed to impact people’s attitudes. Indeed, these treatments not only failed to change people’s attitudes toward peace in the expected direction, but they surprisingly failed to even influence their self-reported mindsets themselves when we use the post-treatment question about wartime mindsets as a manipulation check. Second, we find relatively modest shifts in civilian mindsets between the two waves of the survey. As noted earlier, there was some increase in the survival mindset from wave 1 in July 2022 to wave 2 in February 2023, but the overall sample average as well as the differences across major subgroups remained broadly consistent over time. In particular, the proportion of Ukrainians holding a survival mindset grew from 12 percent in the first wave to 16 percent in the second wave, with similar geographic subsample differences at both times.¹⁶ Our findings imply that, once a large-scale conflict escalates, civilians’ mindsets about it may be slow to move, especially in the absence of significant conflict developments. In sum, the study suggests that civilians’ proclivities toward injustice and survival-oriented mindsets are deeply meaningful—but not easily movable—in wartime.

What are the broader implications of our research? First, the analyses help enrich our understanding of how civilians think—and are likely to act—in war.¹⁷ Indeed, our research suggests that wartime mindsets may help us understand when civilians rally enthusiastically behind one “side” in conflict or simply look for ways to keep their heads down and survive the fighting regardless of the outcome. Crucial wartime behaviors such as volunteering for armed groups may flow from this. Ukraine for example was reportedly overflowing with civilians volunteering for its defense forces and seeking to contribute to the war effort in other ways following Russia’s invasion, especially during the early phases of the fighting. Many of these were likely motivated by a strong injustice-oriented mindset as they looked at the horrors of Russian aggression.

And yet we find ample variation in how Ukrainians understand the conflict. The degree to which those holding an injustice-based mindset transition into holding a survival-oriented mindset—as well as the relative political influence of these two groups—may decide how committed Ukrainians remain to sacrificing their lives and material resources for the war. Sacrifices include their readiness to volunteer for combat, their willingness to endure violence and other material deprivation, and their inclination toward seeking negotiations to bring an end to the conflict. Indeed, key outcomes in conflict studies—such as volunteering to fight and aiding combatants—can be driven by survival when belligerents engage in indiscriminate violence against survival-seekers (Kalyvas and Kocher 2007). In most other conditions, however, we would likely expect survival-seeking civilians to be less likely to fight for either party, less likely to actively aid combatants, and more likely to back peace to hasten the end of a dispute.

Finally, our proposed framework has implications for the bargaining model of war. When the continuation of war is conceived of in terms of a two-player interaction (Filson and Werner 2004; Wagner 2000), one can unpack the utility and costs of those actors as (at least, in part) informed by public resolve to fight and public willingness to support compromise. Our causal logic suggests that a widely prevalent injustice mindset (which is the case in the majority of the regions of Ukraine) in the population would increase a leader’s perceived utility from victory, whereas a widely prevalent survival outlook (which is the case in the frontline regions of the country) should increase a leader’s perceived costs from continued fighting, widening the range of acceptable peace settlements.

Author Biography

Austin J. Knuppe is an Assistant Professor of Political Science at Utah State University, [email protected]. He is the author of Surviving the Islamic State: Cooperation, Contention, and Neutrality in Wartime Iraq (Columbia University Press, 2024).

Anna O. Pechenkina is an Associate Professor of Political Science at Utah State University. Her research examines how peace emerges from war and why it succeeds or fails.

Daniel M. Silverman is an Assistant Professor of Political Science at the Carnegie Mellon Institute for Strategy & Technology (CMIST) at Carnegie Mellon University. His research focuses on international security, political psychology, and the politics of the Middle East and the wider Islamic world.

Notes

Editorial Note: Anna O. Pechenkina placed this manuscript under review, and received the decision to revise and resubmit, before joining the proposal to serve as an Associate Editor for the University at Buffalo-Kansas State ISQ editorial team. This manuscript’s review process was handled by the previous editorial team based at the University of Tennessee throughout the entirety of the review process.

Author’s Note: Names appear alphabetically, authorship is equal. The data underlying this article are available on the ISQ Dataverse, at https://dataverse.harvard.edu/dataverse/isq. The authors thank the following colleagues for their valuable feedback: Jacob Aronson, Celeste Beesley, Greg Goelzhauser, Jia Li, Evan Perkoski, Jeremy Pressman, Diego Romero, Michael Rubin, Josh Ryan, as well as participants in the 2023 West Point Security Seminar and 2022 Four Corners Conflict Network annual conference.

Footnotes

References