Hostname: page-component-cc8bf7c57-fxdwj Total loading time: 0 Render date: 2024-12-10T21:48:53.447Z Has data issue: false hasContentIssue false

Civilized Barbarism: What We Miss When We Ignore Colonial Violence

Published online by Cambridge University Press:  16 October 2023

Paul K. MacDonald*
Affiliation:
Department of Political Science, Wellesley College, MA, USA

Abstract

Colonial warfare has been a frequent and bloody feature of international relations, yet most studies of wartime civilian victimization focus on either interstate or civil wars. In this article I argue that ignoring colonial violence has distorted our understanding of state-directed violence against civilians in wartime. I introduce a new theory of colonial violence, which focuses on the distinctive strategic, normative, and institutional incentives that colonial powers have to harm civilians. To assess this theory, I introduce and analyze a new data set of 193 cases of colonial war from 1816 to 2003. Using a variety of measures of civilian harm, I find that colonial wars are especially brutal. Three-quarters of states in colonial wars targeted civilians, for example, compared to less than a third of states in interstate wars. But some colonial wars are harder on civilians than others. Colonial powers are more likely to harm civilians when their indigenous adversaries employ guerrilla tactics, when their indigenous adversaries come from a different perceived racial background, and when the colonial state relies on settlers or indigenous intermediaries to help compensate for its relative weakness. By ignoring colonial violence in world politics, we misunderstand the scale and scope of state-directed violence against civilians and miss an opportunity to deepen our understanding of the causes of this brutality.

Type
Research Article
Copyright
Copyright © The Author(s), 2023. Published by Cambridge University Press on behalf of The IO Foundation

Colonial warfare has been a frequent and bloody feature of international politics. In the early nineteenth century, revolutions in Spanish America killed tens of thousands of civilians, with royalist armies “sweeping into towns and pillaging, raping, and murdering the occupants into submission.”Footnote 1 During the 1857 Sepoy Rebellion, British forces stormed Delhi, “sacking and looting the Mughal capital and massacring great swaths of the population.”Footnote 2 At the turn of the twentieth century, the Dutch military launched a series of offensives across the Indonesian archipelago during which the “population was punished arbitrarily, severely, and collectively.”Footnote 3 The interwar period featured uprisings across the Middle East, including one in Syria that the French suppressed by bombarding Damascus by means of artillery, tanks, and airplanes over multiple days, killing thousands and leaving “an entire quarter of the city … a smoldering ruin.”Footnote 4 Estimating the human toll from these conflicts is difficult, but according to one study, wars of colonial expansion alone resulted in the deaths of 25 to 30 million people, 95 percent of them civilians.Footnote 5

Despite the bloody character of colonial wars, most studies of wartime civilian harm focus on either interstate or civil wars. For interstate conflicts, scholars have explored the extent to which regime type, the identity of the combatants, the character of warfare, wartime objectives, and international legal obligations influence decisions to target civilians.Footnote 6 Scholars of civil wars have examined how the structure of the government and its armed forces, the ideological and organizational character of rebel groups, and the international context can shape when parties harm civilians.Footnote 7 The only study that has examined civilian harm in the context of colonial wars is that of Valentino, Huth, and Balch-Lindsay, who pool interstate, civil, and colonial wars in their analysis of mass killings after World War II.Footnote 8 Yet they do not consider the distinct mechanisms driving civilian targeting in colonial settings.

In this article, I argue that ignoring colonial warfare has impoverished our understanding of state-directed violence against civilians. Most studies of civilian harm in interstate wars have focused on strategic factors, emphasizing how a state's war aims and wartime strategy can create opportunities and incentives to target noncombatants.Footnote 9 These same studies have downplayed the role of ideational factors, finding a contested relationship between combatant identity, regime type, international legal commitments, and civilian harm. This emphasis on strategic factors makes sense, given that these wars are fought between sovereign states using organized militaries to advance their national interests.

But colonial wars differ from interstate wars in fundamental ways. They are fought against nonstate adversaries, outside a state's recognized international boundaries, with the aim of either establishing or sustaining hierarchic relations of imperial rule. I argue that this setting creates distinctive strategic, normative, and institutional incentives for colonial powers to harm civilians. Colonial militaries are often operating against adversaries who exploit their mobility and knowledge of the terrain to avoid battle, which can create incentives for colonial powers to embrace “scorched earth” tactics. Colonial powers are often confronted by adversaries who they perceive as racially inferior, which can make it easier for them to justify punitive collective punishments. The colonial state often struggles to generate the authority required to effectively govern, so it outsources responsibility for repression to intermediaries who apply force indiscriminately. The presence of settlers can further exacerbate these dynamics by hardening racial attitudes and unleashing escalating violent competitions over land and labor that can become “eliminationist,” if not genocidal, in character.

To assess these possibilities, I introduce and analyze a new data set of 193 cases of colonial warfare from 1816 to 2003. Using a variety of measures of civilian harm, I find that colonial wars are especially brutal. Three-quarters of states in colonial wars targeted civilians, compared to a third of states in interstate wars. Mass killing occurred nearly twice as frequently in colonial wars as in interstate wars. Colonial militaries would burn down villages, destroy crops, and seize livestock. Colonial security forces would conduct mass arrests, employ torture, and execute suspects based on little evidence. In short, the empirical record confirms that colonial conflicts deserve their reputation as “dirty wars.”

But some colonial wars are harder on civilians than others. The data suggest that colonial powers are more likely to target civilians when their indigenous adversaries employ guerrilla tactics, when their indigenous adversaries come from a different racial background, and when the colonial state relies on settlers or indigenous intermediaries to compensate for its frail governance capacity. When we ignore colonial wars, therefore, we underestimate the symbolic role that violence can play in certain settings, especially as a means of reinforcing racial hierarchies and protecting settler privileges. In interstate wars civilian harm is relatively rare and is adopted as a tactic of last resort, while in colonial wars it is commonplace and is embraced as a necessary tool of performative racialized retribution.

The rest of this article proceeds as follows. The first main section describes how existing studies have neglected colonial violence and offers some possible explanations for this oversight. The second section develops a theory of colonial violence, focusing on its unique strategic, ideational, and institutional features. The third section introduces the colonial war data set and sketches out some of the broad features of colonial violence over the past two centuries. The fourth section examines the correlates of civilian harm in colonial wars. The final section offers some observations about the benefits of integrating colonial violence into the study of wartime civilian targeting.

The Neglect of Colonial Violence

Colonial violence has not been a major focus of the study of conflict. To some extent this is an accident. As it refined its data sets, the Correlates of War (COW) project drew a distinction between three kinds of war: “interstate” wars, fought between recognized member states of the international system; “intrastate” or civil wars, fought between a state and nonstate actor within a state's territorial boundaries; and “extra-state” or colonial and imperial wars, fought between a state and a nonstate actor outside the state's territorial boundaries.Footnote 10 An unintended consequence of partitioning the data in this manner was the neglect of colonial violence in quantitative research. As Sarkees, Wayman, and Singer lament, “In contrast to virtually hundreds of publications related to the [COW] inter-state war data set, hardly any scholarship has examined the extra-state war data set.”Footnote 11

There have been studies of the causes of conflict that pooled data on interstate and colonial wars.Footnote 12 There have also been studies of civil war onset that include post-1945 anticolonial wars in their analyses.Footnote 13 Colonial wars also feature prominently in certain specialized data sets. Lyall and Wilson's data set on counterinsurgency wars contains numerous colonial cases,Footnote 14 although they do not center this in their analysis of declining trends in incumbent performance.Footnote 15 Arreguín-Toft includes a number of colonial wars in his data set on “asymmetric conflict,” but he intermixes these cases with asymmetric interstate and civil wars.Footnote 16 In short, there have been relatively few quantitative studies that have examined colonial violence on its own terms,Footnote 17 and none that has explored how the violence states unleash in colonial wars might differ in intensity or kind from violence in other conflicts.

The neglect of colonial violence is not just an issue for quantitative research. Most qualitative studies of war focus on interstate wars, especially those involving European great powers.Footnote 18 Those that focus on the “periphery” of the system tend to emphasize interstate rivalries and territorial scrambles, rather than colonial violence.Footnote 19 Perhaps most surprisingly, even studies of overseas empires can sometimes downplay the extent of colonial violence. In his history of the British Empire, Ferguson neglects to mention the violence that accompanied the suppression of postwar revolts in Kenya and Malaya, in which upwards of 30,000 civilians were killed.Footnote 20 In his survey of American involvement in “small wars,” Boot claims that the American occupation of the Philippines was “less brutal than some critics have charged,”Footnote 21 yet he omits discussion of the massacre at Bud Dajo, in which six hundred Moros, including women and children, were slaughtered.Footnote 22

There are many reasons why even these dramatic examples of colonial violence are minimized. One is that the colonial powers went to great lengths to hide their brutality. In the interwar period, British military writers articulated a doctrine of imperial policing centered on “minimum force,” an idea that supposedly reflected British common law traditions and norms of “gentlemanly restraint.”Footnote 23 Yet in practice, “coercion through exemplary force” remained “the mainstay of British counterinsurgency policy” through decolonization.Footnote 24 A related reason is that colonial officials would employ euphemisms to describe colonial violence. Massacres were redescribed as “battles.” Prisoners were not summarily executed but “shot while trying to escape.” These rhetorical deflections were not accidental but designed for metropolitan audiences, who would get “upset a little too quickly,” in the words of one French official, when confronted by reports of colonial brutalities.Footnote 25

A final reason colonial violence is ignored is that the testimonies of colonized peoples, who experienced the violence firsthand, are rarely collected or considered. The sources that we do have speak to the traumatic character of colonial wars. A pictographic account of the 1876 Battle of Greasy Grass (Little Bighorn), drawn by the Minneconjou warrior Red Horse, shows “the battle's claustrophobic messiness and carnage.”Footnote 26 An oral history provided by a witness to the 1906 Dutch assault on the Balinese kingdom of Badung, which ended in the slaughter of over a thousand civilians, recounts how “modern weaponry met weapons of yore … blood spurted and bodies piled higher.”Footnote 27 A 1914 poem by an anonymous Amazigh performer describes a French raid in the Atlas Mountains, in which “The French use fire / It burns the land / Fire stirs the wind from the west / It kills everything that it touches.”Footnote 28

Whatever the reasons, the neglect of colonial violence is unjustifiable given the importance of colonial empires. At their height prior to the Second World War, colonial empires “spread over 42 percent of the planet's area and affected 32 percent of its inhabitants.”Footnote 29 Violence played a central role in the establishment, maintenance, and ultimate collapse of these empires. Moreover, the experience of colonial violence left important legacies, shaping patterns of postcolonial state-building and the prospects of postcolonial violence.Footnote 30 Finally, the neglect of colonial violence is a missed opportunity to enrich our theories of political violence, given diversity in the modes of colonial rule and patterns of resistance across different empires.Footnote 31

A Theory of Colonial Violence

Colonial wars are staggeringly diverse and share features with both interstate and civil wars. Some colonial wars, such as the 1826 Russo–Persian War or the 1839 Anglo–Chinese Opium War, were fought along conventional lines between organized armies, and thus resemble interstate conflicts. Others, such as the 1936 Arab Revolt and the 1954 Algerian War, featured small groups of rebels utilizing guerrilla tactics, and thus anticipate many contemporary civil conflicts. Between these extremes we have numerous examples of indigenous polities opposing colonial empires, with various kinds of military forces using a wide range of tactics.

Given this variety, we must be careful in generalizing about colonial violence. Yet for all their differences, colonial wars do share some characteristics. In particular, colonial wars are (1) fought by states outside their own territorial boundaries, (2) against adversaries who are not recognized members of the international system, (3) with the goal of establishing or sustaining hierarchical relations of dominance. I argue that these features create strategic, normative, and institutional tendencies that influence how colonial wars are fought, and I suggest specific hypotheses about when colonial powers are more likely to target civilians.

Strategic Character of Colonial Wars

One feature of colonial wars is that they are often fought by irregular forces using guerrilla methods. As Walter argues, many indigenous adversaries of empire “avoided engaging in battle wherever possible” and “instead relied upon concealment, ambushes and surprise attacks, on ruses such as feigned retreats, and on surrounding and ‘cutting off’ the enemy.”Footnote 32 Indigenous polities might opt not to fight using conventional methods for a variety of reasons. In some cases they embrace guerrilla methods for tactical reasons, seeking to exploit their greater mobility, knowledge of the terrain, and intelligence advantages.Footnote 33 In other cases they use such methods because they are well suited to indigenous military manpower systems, which are centered on small bands of fighters operating independently.Footnote 34 Indigenous polities may also avoid open battle in response to technological factors, withdrawing to fortifications or rough terrain to nullify colonial advantages in firepower.Footnote 35

As in other conflicts, the use of guerrilla methods by one side can generate strategic incentives for the other side to target civilians.Footnote 36 First, colonial powers may target civilians to deny guerrillas material resources, such as weapons, food, or fresh recruits. Moving civilians into fortified villages is a familiar tactic to interdict supplies intended for rebels.Footnote 37 Second, colonial powers can use civilian targeting as an intimidation tactic, to deter civilians from aiding the rebels. Burning a village suspected of aiding the guerrillas can send a message to neighboring villages.Footnote 38 Third, guerrilla tactics can take on symbolic meanings in colonial settings. Societies that embrace guerrilla modes of warfare are perceived as unchivalrous, and thus less deserving of restraint, than those who fight “honorably” in the open.Footnote 39 Whatever the motive, civilians are easier to target than elusive rebels, so civilian victimization is a cheaper counterinsurgency tactic than the systematic clearing and holding of territory.Footnote 40 All of this suggests:

H1 In colonial wars, states are more likely to target civilians when indigenous adversaries adopt guerrilla tactics.

Ideological Content of Colonial Wars

A second shared feature of colonial wars is that the indigenous adversaries of empire are not recognized members of the international system, and are often perceived as having different and inferior racial identities. The kinds of actors that have taken up arms against empires are varied, and have included military monarchies, modernizing proto-states, and “tribes” or chiefdoms, among many others.Footnote 41 What unites these actors is that colonial powers often perceive them as racial others: rather than accept them as “rational” or “civilized” states, they stigmatize them as “barbaric” or “savage” societies. Of course, the specific ways in which colonial powers understand racial differences can vary from empire to empire, and can shift over time.Footnote 42 Yet as imperial historians have emphasized, by the nineteenth century, the “color line” had become a core principle around which European empires were organized.Footnote 43

Perceptions of racial difference are not just central to the ideology of empire; they can also influence practices of colonial violence. First, they can be used to justify performative collective punishments. Because “backward” peoples purportedly act in impulsive and collective ways, colonial powers find it harder to maintain the distinction between combatants and civilians. They come to see themselves as involved in wars against entire societies, which must be chastised through collective punishments.Footnote 44 Because “barbaric” societies are governed through despotic force, colonial powers likewise assume that they will respond to only spectacular “performances of punitive violence.”Footnote 45 Second, perceptions of racial difference can be used to exempt societies from the usual normative or legal restraints related to war. “Construction of the enemy as ‘un-civilized’,” Wagner argues, “dictated and justified techniques of violence that were … considered unacceptable in conflicts between so-called ‘civilized’ nations.”Footnote 46 Because they are not recognized states, indigenous polities do not fall under the traditional ambit of international law.Footnote 47 And because nonwhite societies are perceived as “fanatical,” colonial powers believe they must adopt “emergency” legal frameworks that allow harsh measures.Footnote 48 All of this suggests:

H2 In colonial wars, states are more likely to target civilians when indigenous adversaries are perceived as having a distinct and inferior racial identity.

Institutional Context of Colonial Wars

A third feature of colonial wars is that they are fought to establish and sustain hierarchical relations of imperial rule. A key feature of imperial hierarchies, like all hierarchies, is that the dominant state (or metropole) is entitled to command, while the subordinate polity (typically a colony) is required to obey.Footnote 49 Yet imperial hierarchies are also distinctive. Metropolitan governments deprive subordinate societies of some, but not necessarily all, of their sovereignty.Footnote 50 They then appoint representatives—typically a governor, high commissioner, or some similar figure—to oversee their interests. In turn, these “men on the spot” must negotiate with indigenous intermediaries—often a king, paramount chief, or similar local authority—to share the responsibility of governing the population. The colonial state, which emerges from these various compromises, is unique. Compared to sovereign states, the colonial state has a limited governance capacity. Its territorial boundaries are often undefined, its bureaucratic infrastructure limited, and its administrative reach circumscribed. In many cases, a “thin white line” of metropolitan officials attempts to govern a large indigenous population, with colonial authority struggling to reach beyond coastal enclaves or capital cities.Footnote 51 “Even during its heyday,” Conrad and Strange conclude, “the colonial state was essentially a weak state.”Footnote 52

Given its weaknesses, the colonial state often becomes heavily dependent on intermediaries, who officials hope will augment its governance capacity and authority. Yet this reliance on intermediaries creates two potential pathways for civilian harm. First, the interests of the intermediaries and the colonial officials overlap but are not the same. Thus the situation is vulnerable to classical principal–agent problems, where intermediaries seek to exploit the authority granted them for their own parochial purposes.Footnote 53 Second, the colonial state is often unwilling or unable to exercise control over its intermediaries, in part because it fears undermining their effectiveness. This trade-off between competence and control becomes particularly acute in wartime, when the colonial state is desperate for resources to fend off armed challenges.Footnote 54 In these situations, the colonial state may let its intermediaries loose, sacrificing control for competence, in ways that allows them to abuse civilians.

While all colonial powers struggle to manage intermediaries, there are important variations between colonies. “In some contexts, colonial rule thickened into an effective apparatus of surveillance and punishment,” Burbank and Cooper observe, “but elsewhere its presence was thin, arbitrary, and episodically brutal.”Footnote 55 I distinguish between three types of colonial institutions. Settler colonies are colonial dependencies where a significant number of settlers reside on a semipermanent basis. These settlers, Robinson argues, are “ideal prefabricated collaborators.”Footnote 56 They can supplement the coercive capacity of the colonial state by volunteering for service in colonial militias. They can act as landlords or employers, integrating indigenous populations into expanding colonial economies. Yet the presence of settlers can also open up pathways of civilian harm. First, settlers tend to have different—and often more extreme—interests than colonial authorities. They tend to see colonial wars as opportunities to loot indigenous wealth, expropriate land, smash indigenous social systems, and force individuals into colonial labor markets.Footnote 57 They also tend to be more invested in the maintenance of colonial racial hierarchies and thus are more willing to sanction punitive policies that restore settler prestige.Footnote 58 Second, settlers often have competencies that colonial authorities are reluctant to sacrifice. Settler militias are seen as particularly adaptable to local modes of warfare. Settler leaders are assumed to have a strong understanding of indigenous populations. In wartime, the colonial state may feel pressure to make concessions to settler interests. All of this suggests:

H3a In colonial wars, states are more likely to target civilians when large populations of settlers are present.

Compare these dynamics to those found in indirect-rule colonies. Here the colonial state relies on indigenous “intermediaries,” “collaborators,” or “loyalists.” These actors can be used to police subject populations, administer justice, allocate land, collect taxes, and recruit corvée labor, among other tasks.Footnote 59 As with settlers, however, the reliance of the colonial state on indigenous intermediaries can put civilians at risk. First, indigenous intermediaries do not have the same interests as colonial authorities. They seek opportunities to exploit the arbitrary authority granted to them by the colonial state to monopolize resources, eliminate rivals, and appropriate the wealth and labor of their followers.Footnote 60 Colonial wars can thus provide opportunities for indigenous intermediaries to further entrench the systems of “decentralized despotism” that Mamdani argues are a core feature of indirect rule.Footnote 61 Second, indigenous intermediaries have competencies that colonial authorities prize during wartime. Indigenous fighters are valued for their knowledge of the terrain, tactical skill, and—in a racist perception—their “ruthlessness.” Indigenous elites are seen as valuable sources of local authority. In wartime, the colonial state will often increase its reliance on indigenous intermediaries, granting them license to use force indiscriminately to restore order. All of this suggests:

H3b In colonial wars, states are more likely to target civilians in indirect-rule colonies.

Compare both of these dynamics to those in direct-rule colonies, where the metropole makes extensive use of its own administrators. Instead of grafting precolonial institutions onto the colonial state, it supplants them with its own bureaucratic structures for raising revenues, administering justice, and policing populations.Footnote 62 This does not free the colonial state from some reliance on intermediaries, but it does channel collaboration into more formal settings, with indigenous actors serving as clerks, translators, and military recruits. Building direct-rule institutions can be expensive and time consuming, of course, and a colonial power's ability to do so can depend on the colony's size, revenue base, and precolonial governance institutions. Yet, as Lange argues, once they are established, direct-rule colonies tend to exhibit greater “infrastructural power.”Footnote 63 Like all colonial states, they maintain control through repression. Yet because they have a more developed coercive capacity, with more robust police forces and more extensive intelligence apparatuses, they are able to anticipate and defuse conflict more quickly.Footnote 64 Moreover, the indigenous intermediaries in direct-rule colonies tend to be different from their indirect-rule counterparts. Because they are more deeply entrenched within the colonial bureaucracy, they tend to have interests that align more closely with the colonial state, reducing principal–agent problems. Given the infrastructural power of direct-rule institutions, colonial officials are able to empower intermediaries without sacrificing control over them. All of this suggests:

H3c In colonial wars, states are less likely to target civilians in direct-rule colonies.

Patterns of Colonial Violence

To test this theory of colonial violence, I compiled a data set of 193 cases of state participation in a colonial war between 1816 and 2003. Following Sarkees, Wayman, and Singer, I define a colonial war as sustained combat between a territorial state and a nonsovereign entity outside the borders of that state that results in at least a thousand combined fatalities over the course of the conflict.Footnote 65 Within this broad category, I confine my analysis to colonial wars between European states or other great powers and nonsovereign entities outside Europe. Thus I do not consider cases of territorial conquest within Europe, such as the partition of Poland. I also do not consider cases of expansion by non-European empires, such as the Chinese subjugation of Tibet. These restrictions allow me to focus on what scholars consider the most essential and puzzling feature of modern imperialism: the dramatic expansion of European states into the “periphery” of the international system over the course of the nineteenth and twentieth centuries.Footnote 66

I identified cases using a variety of sources. First, I included every case of a colonial war in the COW extra-state war data set, excluding those within Europe (such as the Garibaldi Expedition) and those that did not involve a European state or a great power protagonist (such as Egypt's suppression of Sudanese slave traders). I also excluded cases that developed as part of an ongoing interstate war (such as the German East Africa campaign) and cases that were unresolved as of 2003 (such as the US war in Afghanistan). Second, I identified cases of colonial violence in specialized data sets on asymmetric conflict and counterinsurgency warfare that are not included in the COW extra-state war data set.Footnote 67 They included cases of violence across frontiers against indigenous peoples by expanding continental states, such as the Sioux Wars in the United States. Finally, I consulted various specialized encyclopedias of both warfare and colonialism.Footnote 68 From these sources, I identified an additional eighteen cases that had not appeared in prior data sets, including the Sétif uprising in postwar Algeria. A complete list of cases can be found in Appendix 1.

Figure 1 plots the frequency of the onset of colonial and interstate wars in each decade between 1816 and 2003. In general, colonial wars have been more common than interstate conflicts, with an average of 9.3 colonial wars starting per decade compared to 5.0 interstate wars. Outbreaks of colonial violence peaked in the late nineteenth century during the territorial scrambles associated with the “new imperialism.” They subsequently declined over the twentieth century, before nearly disappearing following the postwar wave of decolonization. A consideration of colonial violence, however, reframes how scholars periodize eras of war and peace. It is possible to talk about the nineteenth century as a period of “stability” only if we exclude colonial violence.Footnote 69 Claims of a “long peace” following the Second World War likewise require us to ignore the significant bloodshed that occurred in various wars of national liberation.Footnote 70

Figure 1. Frequency of Colonial and Interstate Wars, 1816–2003

Coding Civilian Harm in Colonial Wars

My primary outcome of interest is the intensity of state violence in colonial wars. To facilitate comparisons with existing studies, I collected data around three wartime practices. First, I examined whether colonial powers engaged in civilian victimization, which Downes defines as “a wartime strategy that targets and kills (or attempts to kill) noncombatants.”Footnote 71 Common examples of civilian victimization in colonial wars include the systematic burning of villages and the widespread destruction of crops. One challenge in coding civilian victimization is the dissonance between official policy statements and actual practice. I therefore code civilian victimization in terms of the actual battlefield practices sanctioned by leaders of colonial security forces, rather than the declarations of metropolitan officials or colonial governors.

Second, I catalog the degree of brutality employed by the colonial state across four issue areas: (1) its treatment of civilians, (2) its treatment of prisoners, (3) its use of “inhumane” weapons, and (4) its use of aerial bombardment. Following Morrow and Jo, who undertake a similar coding exercise for interstate wars,Footnote 72 I score the colonial state's methods on a four-point ordinal scale: (1) no brutality reported; (2) only minor cases of brutality; (3) major cases of brutality occur, but the state makes some attempt to minimize harm; and (4) major cases of brutality occur frequently and without constraint. For treatment of civilians, I record whether the colonial state engaged in the shelling of population centers, village burning, food control, collective punishments (such as fines and embargoes), forced resettlement, indiscriminate massacres, systematic looting, and widespread sexual violence.Footnote 73 For treatment of prisoners, I observe whether the colonial state employed mass arrests, summary executions, show trials, torture, deportation, forced labor, and the mutilation and display of dead bodies. For inhumane weapons, I note whether the colonial state used weapons that were considered inhumane by contemporaries, including chemical weapons, expanding “dum dum” bullets, and napalm and other defoliants. For aerial bombardment, I look for examples where the colonial state used fixed- or rotary-wing aircraft to target civilians, their homes, or their food supplies.

Third, I examine whether the colonial state engaged in mass killing, which Valentino defines as the intentional killing of at least 50,000 noncombatants over the course of five years or less.Footnote 74 Accurate data on civilian deaths are difficult to collect in almost all conflict environments, but matters are even more complicated in the colonial context. Colonial governments tend to exaggerate the battlefield losses of their adversaries, trumping up indecisive skirmishes into decisive victories, while downplaying the suffering of civilians. Imperial historians compensate by using demographic data to generate mortality estimates, but the quality of colonial record keeping varies, and estimates can have large margins of error. For example, Blacker concludes that “excess deaths” during the Mau Mau rebellion were probably not as high as the 300,000 figure cited in some sources, but provides a wide-ranging estimate of 30,000 to 60,000.Footnote 75 I err on the side of caution and require clear evidence of more than 50,000 deaths.

To code individual cases, I consulted a wide variety of sources. Histories of particular empires provided useful background but rarely delved into battlefield practices. Military histories proved more useful, although campaign narratives were less helpful than those that quoted extensively from soldiers’ letters and diaries. I also consulted primary documents, mostly accounts by officers involved in colonial campaigns. Every effort was made to collect narratives from colonized peoples, and when this was not possible, I endeavored to “read against the grain,” as Guha suggests, to identify potential elisions in colonial sources.Footnote 76 All told, the coding materials cite roughly 600 sources in six languages. A complete coding rubric, with examples, can be found in Appendix 2. Despite every attempt to be as thorough as possible, however, it is likely that more taboo behaviors, such as torture and sexual violence, remain underreported in the data.

Comparing Civilian Harm in Colonial Versus Interstate Wars

I begin by comparing state-directed civilian harm in colonial wars with comparable data on interstate wars.Footnote 77 I exclude civil wars largely on the grounds of data availability; the data we have on civilian harm in civil wars either pool government and rebel violenceFootnote 78 or cover only the post–Cold War period.Footnote 79 The overall picture provided by the descriptive statistics is stark: regardless of what metric one chooses, colonial wars are particularly hard on civilians (Table 1). Whereas 31 percent of states in interstate wars engage in civilian victimization, 76 percent of state participants in colonial wars do so.Footnote 80 These findings contradict Downes, who finds (using a smaller sample of eighty-four cases of COW extra-state wars) that only 29 percent of colonial powers targeted civilians.Footnote 81 I suspect this discrepancy is driven by two factors. First, my data set is based on the updated COW extra-war data set, which provides a larger and more complete sample of cases. Second, I had the opportunity to consult more recent historical work, which has uncovered significant abuses in canonical cases, such as the 1936 Arab revolt, which had previously been viewed as models of restrained warfighting.Footnote 82

Table 1. Comparing Violence in Colonial Versus Interstate Wars, 1816–2003

Note: Numbers of cases in parentheses.

Colonial wars are also fought using more brutal methods than interstate wars. This is particularly true if we compare the behavior of colonial powers with Morrow's coding of state compliance with the laws of war. Using a four-point scale from full compliance to noncompliance, colonial powers are significantly more likely than states in interstate wars to mistreat civilians (3.5 versus 3.0), to abuse prisoners (3.3 versus 2.8), and to bomb civilians (2.5 versus 2.1).Footnote 83 The one exception to this finding concerns the use of inhumane weapons, where colonial powers and states are roughly equal in compliance (1.5 versus 1.4).Footnote 84 The most commonly reported civilian harm in colonial wars is village burning (72 percent of cases), followed by food destruction (45 percent), bombardment of towns (36 percent), and forced resettlement (31 percent). The most frequent kind of prisoner mistreatment is summary execution (55 percent of cases), followed by torture (23 percent), mass arrests (21 percent), and deportation (21 percent).

Colonial wars are also more likely than interstate wars to feature mass killings. While 6 percent of states engaged in mass killing in interstate wars, more than 12 percent of states did so in colonial wars.Footnote 85 Again, my data report a higher rate of colonial mass killings than previous studies. Valentino, Huth, and Balch-Lindsay identify two cases of mass killing in post-1945 extra-state wars (Indochina and Algeria), but my data set has eight such cases.Footnote 86 This discrepancy probably stems from my use of more recent historical work. Dutch brutality in Indonesia, for example, has been the subject of significant revisionist scholarship, which has found that “atrocities and clear cases of war crimes occurred on a regular basis and may even have been systematic.”Footnote 87 Scholars have also revisited French repression in Madagascar, which was the site of “the worst violence in a French African territory since the Rif War in Morocco twenty years earlier.”Footnote 88 The emerging picture is that the intense violence that accompanied the end of empire in well-known cases such as Algeria was far from an exception and more the rule.

Correlates of Colonial Violence

Colonial violence is worth examining not only because of its distinctive character but also because it provides a useful setting in which to test existing theories of wartime civilian targeting. As I noted in the introduction, most studies focus on the strategic incentives states have to target civilians and downplay the importance of normative factors. In what follows, I use the colonial war data set to test whether these findings hold when we examine civilian victimization in colonial settings.

Research Design

My unit of analysis is the state participant in a colonial war.Footnote 89 My dependent variable is civilian victimization, as defined by Downes.Footnote 90 Drawing on my theory of colonial violence, I consider three core explanatory variables. The first is the strategic character of the war, in particular whether an indigenous adversary employs guerrilla tactics. Coding this variable can be complicated. Ferris notes that indigenous forces in colonial wars frequently adopt “hybrid” methods, employing a mix of “regular and irregular forces” that would use their “conventional weapons in unconventional ways.”Footnote 91 Colonial wars can also change over time: the 1899 Boer War opened with a series of conventional battles and sieges, before evolving into a prolonged guerrilla conflict. I apply relatively stringent criteria for the category of a guerrilla war. Evasion and harassment must be the primary approach of an indigenous adversary throughout the conflict. I code wartime strategy as 1 if guerrilla methods predominate, and 0 otherwise.

The second explanatory variable relates to combatant identity, in particular whether colonial powers perceive their indigenous adversaries as having a different racial identity. Because race is a socially constructed concept whose meaning is constantly changing, coding this variable can be fraught.Footnote 92 Race can also take on multifaceted meanings in colonial settings. Officials in British India considered themselves superior to their colonial subjects, but also believed in a distinction between “martial” and “non-martial” races.Footnote 93 British military officers perceived their opponents in the Boer War as white, but also denigrated them as “backward” and “corrupt.”Footnote 94 To simplify matters, I follow theorists such as DuBois and focus on the overriding importance of the “color line” in colonial settings.Footnote 95 “The colonized world,” Fanon observes, “is a world divided in two,” where colonial powers draw sharp distinctions between white and nonwhite peoples and places.Footnote 96 I draw on the Ethnic Power Relations data set and its coding of a group's socially constructed racial marker, based on its origin in one of seven world regions.Footnote 97 I code combatant identity as 1 if the indigenous adversaries’ racial identity differs from the colonial power, and 0 otherwise.

The third set of explanatory variables concerns the structure of the colonial state, in particular whether a colony is a settler colony or is under indirect rule. Coding these variables is also challenging. Despite a rich literature on settler colonialism, there is no commonly accepted definition of a settler colony.Footnote 98 There are always a smattering of individuals from the metropole present in colonial settings, such as soldiers, traders, or missionaries. What makes settler colonies unique is that these people settle in large numbers, more or less permanently. The most comprehensive data on settlers come from Easterly and Levine, who code the share of European population in a given colony during its formative years.Footnote 99 I code the settler colonialism variable as 1 if the share of settlers exceeded 0.5 percent of the total population, and 0 otherwise.

Similarly, there is no consistent cross-colonial measure of indirect rule. Scholars have used various indicators—such as the proportion of court cases handled by native courts, the relative density of colonial road networks, or the survival rate of precolonial political dynasties—as proxies, yet these indicators are not available for all empires, colonies, or periods.Footnote 100 Further complicating matters is the fact that different regions within a colony can have different forms of rule. British India featured a mix of direct and indirect rule, depending on whether a territory was ruled by a native prince. At the risk of oversimplification, I consider indirect-rule colonies to be those where colonial officials are heavily dependent on indigenous institutions. For British colonies, I draw on Lange's data and consider colonies where more than half of court cases are handled by customary courts to be under indirect rule.Footnote 101 For non-British colonies, I rely on administrative histories to assess the degree of colonial dependence on local collaborators. The resulting variable is coded as 1 when a colony is governed indirectly, and 0 otherwise.

In addition to these core explanatory variables, most of my models include a battery of control variables, which I derive from the existing literature.Footnote 102 These control variables include measures of the colonial power's war aims, the colonial power's regime type, the colonial power's aggregate military capabilities, the extent to which a colonial power has ratified international legal covenants related to the laws of war, the degree of professionalism in a colonial power's military, and the duration of the colonial war. A complete description can be found in Appendix 3. In addition to these controls, some models also include fixed effects for colonial power, region, or both. Some scholars have claimed that the French style of colonial warfare, as developed by Galliéni, Lyautey, and others, favored “peaceful penetration” rather than punitive methods.Footnote 103 Colonial power fixed effects take into account the possibility that these kinds of unmodeled differences between colonial powers might account for their targeting choices. Other scholars have speculated that colonial wars fought in sub-Saharan Africa impose particular challenges on combatants due to the continent's climate, geography, and disease profile.Footnote 104 Region fixed effects take into account the possibility that these kinds of unmodeled regional differences might influence patterns of colonial warfare.

Empirical Results

Because my dependent variable is dichotomous, all models are estimated using a logit model, with robust standard errors clustered by conflict. Table 2 reports the results of five models. Model 1 includes only my core explanatory variables. Model 2 adds the control variables. Model 3 includes the controls plus colonial power fixed effects. Model 4 includes the controls plus region fixed effects. Model 5 considers the controls and both colonial power and region fixed effects. The complete regression tables for all five models can be found in Appendix 4. I discuss each of my core explanatory variables in turn.

Table 2. Correlates of Civilian Victimization in Colonial Wars

Notes: Robust standard errors in parentheses. *p < .05; **p < .01; ***p < .001.

Civilian Victimization in Guerrilla Colonial Wars

The coefficients for wartime strategy are positive and statistically significant at the p < .01 level across all five models, providing strong support for Hypothesis 1. The predicted probability that a colonial power will target civilians in a colonial war is eighteen percentage points higher when an adversary adopts guerrilla tactics (Figure 2a).Footnote 105 As in both interstate and civil wars, wartime strategy appears to be strongly associated with decisions to target civilians.Footnote 106 In this important respect, colonial wars have much in common with other kinds of conflicts.

Figure 2. Marginal Effects of Core Explanatory Variables

While the models suggest a strong association between guerrilla colonial wars and civilian victimization, the case narratives provide evidence for the particular causal mechanisms. First, colonial powers routinely targeted civilians to deny guerrillas access to resources. During the 1911 Fez Revolt, French officials decided that “to reduce such a tenacious enemy, it is necessary to burn their crops without mercy … Only then, will they come to settle.”Footnote 107 Portuguese officials reached a similar conclusion during the 1962 Guinea-Bissau War, endorsing the napalming of rural villages to “starve [insurgent] forces to death or better still force them to surrender.”Footnote 108 British authorities during the 1948 Malaya Emergency viewed the forcible resettlement of nearly 500,000 people in similar terms, stating that “if these areas are dominated to such an extent that food, money, information and propaganda were denied the enemy … the initiative then becomes ours.”Footnote 109

Second, colonial powers frequently targeted civilians to deter them from aiding guerrillas. During the 1816 Caucus Campaign, Russian commanders argued that shelling villages was “necessary as an example … to other peoples, who can be tamed only through the lessons of terror.”Footnote 110 French officials justified the bombing of suspicious hamlets during the 1930 Yen Bay uprising because “it was important to inflict on the bandits and those sheltering them a quick and exemplary lesson.”Footnote 111 Spanish authorities employed a similar logic during the 1821 Venezuelan Revolution, when they executed civilians suspected of harboring republican sympathies and displayed their cadavers “to terrorize noncombatants into either submission and loyalty or flight and exit.”Footnote 112

Third, colonial powers developed pejorative images of adversaries who adopted guerrilla tactics, as a pair of wars fought by the British East Indian Company illustrate. In 1855, the Santal tribe in Bengal fought a grueling six-month guerrilla campaign in which insurgents burnt rival villages, plundered indigo factories and railway works, and harassed colonial security forces.Footnote 113 Company officials concluded these guerrilla tactics required a harsh response. “Forbearance towards ravaging and exterminating bands perpetrating unlimited and unparalleled atrocities, is cruelty to the multitude of victims,” the Board of Control concluded, endorsing the imposition of martial law.Footnote 114 Governor General Dalhousie likewise approved harsh tactics—including the systematic destruction of Santal villages—stating simply that “these people have ceased to deserve mercy.”Footnote 115 Compare this to the Company's reaction to the 1846 Sikh War, which was largely fought along conventional lines. In this case, British sources are full of praise for the Sikhs and their martial spirit. One general wrote of the “rare species of courage possessed by these men.”Footnote 116 Dalhousie acknowledged that “the Sikhs behaved bravely, and stood their ground obstinately.”Footnote 117 Given the nature of the fighting, the Company did not engage in systematic village burning or crop destruction; instead, set-piece battles were followed by rounds of prisoner exchanges and diplomatic negotiations. Conventional war provided neither the incentive, nor justification, for brutality.

Civilian Victimization in Racialized Colonial Wars

Turning to the role of race, the coefficients for combatant identity are positive and statistically significant at the p < .05 level across four of the five models, providing mixed but positive support for Hypothesis 2. The predicted probability that a colonial power will target civilians in a colonial war is forty-six percentage points higher when their adversaries have racial identities different from their own (Figure 2b). Colonial wars are more likely than interstate wars to be fought by combatants with different racial identities (89 percent versus 57 percent) and feature higher rates of civilian victimization (76 percent versus 31 percent), which provides prima facie evidence that race matters. Yet these results suggest that there is important variation within the category of colonial wars. Colonial wars fought across racial lines are more likely to feature civilian victimization than nonracialized ones, such as when colonial powers are fighting rival white settlers. Recent work by Fazal and Greene finds that in interstate wars European states are more likely to target civilians when fighting non-European opponents.Footnote 118 These results suggest this is generally the case in colonial settings, too.

The case narratives again provide illustrations of the causal mechanisms connecting racial difference to civilian targeting. First, colonial powers frequently drew on race to legitimate collective punishments. In the 1876 Lakota War, American officers justified the shelling of Lakota villages by claiming that “Indians are the most clannish people in existence.”Footnote 119 British officials along the Northwest Frontier of India defended the razing of entire villages by observing that the Pashtun “does not possess … innocent subjects to be spared … All of them to a man [are] concerned in hostilities.”Footnote 120 French officials endorsed harsh methods during the 1894 campaign against Samory Touré based on a similarly racist view that “all the blacks understand is fear.”Footnote 121

Second, colonial powers repeatedly appealed to race to explain that normative restraints on war did not apply in colonial conflicts. German commanders responsible for 1903 genocide of the Herero and Nama were blunt in their assessment that “against ‘nonhumans’, one cannot conduct war ‘humanely’.”Footnote 122 Dutch hardliners responsible for the pacification of Aceh in the 1890s insisted that international law applied to only “European situations” and not to a conflict with “an uncivilized nation.”Footnote 123 Despite its doctrine of “minimal necessary force,” the British army claimed that the “degree of force necessary … will obviously differ very greatly between the United Kingdom and places overseas,” a carve-out that was used early on during the 1952 Mau Mau rebellion to justify mass evictions and extrajudicial killings.Footnote 124

In cases where racial differences were absent, colonial powers found it harder to endorse civilian targeting. During the 1880 Transvaal Rebellion, British generals denigrated Boers as “the most ignorant … of white men,”Footnote 125 yet British forces did not target Boer civilians, and the Cabinet ultimately agreed to restore Boer independence because it feared the conflict might “excite a war of races throughout South Africa.”Footnote 126 During the 1954 Cyprus Emergency, British officials considered the option of collective punishments, but ultimately concluded—in racially coded language—that it was “inappropriate to use such a tribal method against a more developed people.”Footnote 127

A skeptic might point to the 1899 Boer War, in which the British military employed various “methods of barbarism” against Boer civilians—including concentration camps in which thousands died from disease and malnutrition—as a notable exception.Footnote 128 Yet comparing this case to similar wars against nonwhite opponents during the same period supports arguments about the centrality of race. The concentration camps were established in part for humanitarian reasons, to save homeless Boer women and children from starvation, and when liberal critics raised concerns about their horrendous conditions, officials made efforts to improve them.Footnote 129 The British Cabinet also prohibited imperial forces from using “dum dum” bullets, in part because of concerns it would contradict a Hague Conference prohibition.Footnote 130 The British military urged its soldiers to treat detainees humanely, and in one notable case, brought charges against three officers accused of murdering Boer prisoners and civilians.Footnote 131

Compare this to the 1906 Zulu Rebellion, where colonial forces did nothing to help starving African women and children but instead forced them back into the bush;Footnote 132 where colonial forces were equipped with expanding bullets and nobody raised any objections;Footnote 133 and where colonial forces made no effort to protect prisoners. Indeed, two of the rebellion's decisive “battles” were little more than massacres, in which colonial forces slaughtered every African male they encountered.Footnote 134 The point here is not to suggest the Boer War was fought in a humane manner or to minimize anti-Boer chauvinism. Yet because the war was fought between two white opponents, there was at least some belief that violence should be kept under control and at least some effort to use regular parliamentary and legal instruments to do so. No similar expectations or protections were extended to the black Africans who rebelled four years later.

Colonial Institutions and Civilian Victimization

Turning to colonial institutions, the coefficients for both settler and indirect-rule colonies are positive and statistically significant at the p < .05 level across all five models, providing consistent support for Hypothesis 3. The predicted probability of civilian victimization is sixty-one percentage points higher in settler colonies and thirty-four percentage points higher in indirect-rule colonies, relative to a direct-rule baseline (Figure 2c, 2d). Studies by Stanton and Balcells have highlighted the important role that preexisting political institutions can play in shaping decisions to target civilians in civil wars.Footnote 135 Governance institutions appear to play an analogous role in colonial wars, creating opportunities and incentives for local actors, whether settler militias or indigenous auxiliaries, to target civilians.

The case narratives confirm that settler militias were among the most frequent abusers of civilians. During the 1850 Mlanjeni War, settler militias on the Eastern Cape frontier conducted “a brutal campaign of extermination waged against men, women, and children.”Footnote 136 Settler militias unleashed a similar “war of extermination” during the 1860 Apache War, with “Indian hunters volunteering to be paid for scalps.”Footnote 137 In the 1945 Sétif Uprising, settlers formed “self-defense groups” which fanned out into Muslim towns, conducting hastily assembled tribunals. In Guelma alone, these vigilantes executed over a quarter of the town's adult males.Footnote 138

In some cases, settlers targeted civilians to advance material goals. During the Kalkadoon Wars in Queensland, armed groups of white farmers would harass Aboriginal bands, poisoning crops and kidnapping women and children, to clear grazing pasturage.Footnote 139 Settlers seized similar opportunities for gain during the 1893 Matabele War, when raiding parties stole an estimated 100,000 cattle.Footnote 140 In other cases, settlers abused civilians because of their racist views. During the 1832 Black Hawk War, settlers conducted a brutal campaign of retribution, slaughtering women and children, with one militiaman explaining that if you “kill the nits … you'll have no lice.”Footnote 141 Even as norms of racial equality gained strength during the twentieth century, many settlers continued to harbor deep prejudices. During the 1952 Mau Mau rebellion, the settler-dominated Kenya Regiment committed numerous atrocities, an unsurprising outcome given its motto: “The only good Kikuyu is a dead one.”Footnote 142

The case narratives suggest that indigenous intermediaries played a similar role in indirect-rule colonies. During the 1892 Arab War in the Congo, Belgian officials compensated for their relatively thin administrative presence by recruiting various auxiliaries, including cannibals, who would “carry out much of the ‘dirty work’ during the campaign.”Footnote 143 French officials were so reliant on local militias in their 1899 Chad campaign that the pillaging and enslavement of the residents of defeated villages by these intermediaries became an “integral part of French strategy.”Footnote 144 The British were similarly dependent on local levies to police northern Nigeria, and when the town of Satiru rebelled in 1906, British officials sat idly by as these forces massacred its residents.Footnote 145

Colonial powers were reluctant to give up the “competencies” these intermediaries were purportedly providing. During the 1900 War of the Golden Stool, British officials raised local levies called “locusts” from traditional enemies of the Ashanti, assuming that these groups would be highly motivated fighters, and stood by when they would “murder, rape, and … enslave any Asante women and children they were able to capture.”Footnote 146 During the 1898 Philippine insurgency, the American army compensated for its lack of ties to a “class of Filipinos … willing to cooperate” by encouraging its indigenous scout units to torture civilians for information.Footnote 147 The British army addressed its intelligence deficits during the 1936 Arab Revolt by forming “special night squads,” consisting of a mix of British officers and local Jewish fighters, who developed a reputation for “vindictiveness … [and] killing in cold blood.”Footnote 148

In direct-rule colonies, in contrast, colonial officials were able to exert more control over security forces and thus temper their abuses. During the 1921 Moplah Revolt, the government of India mobilized four Indian army regiments and declared martial law, but also took steps to ensure that “petty persecution of inhabitants in places occupied by troops” was “rigorously forbidden.”Footnote 149 Similarly, during the 1942 Quit India movement, colonial authorities recognized that while “maintaining internal order was of utmost importance,” there was a “need to avoid excesses in doing so.”Footnote 150 Officials in provinces such as Uttar Pradesh thus focused on expanding the strength of the civil police and armed constabulary, and called the military out of its barracks only in exceptional circumstances. The sheer scale and complexity of the Raj's coercive infrastructure limited the need to outsource barbarism to unreliable intermediaries.

Robustness Checks and Extensions

To further explore the robustness of these findings, I conducted a series of sensitivity analyses, which are reported in full in Appendix 5. I consider various ways of operationalizing key variables, yet substituting alternative measures for wartime strategy, military professionalism, democracy, and international legal obligations do not alter the main results. I explore whether the results are biased by the fact that Great Britain is the incumbent state in almost 40 percent of colonial wars, yet the results are unchanged when I include a dummy variable for Britain's imperial wars. I assess whether the inclusion of wars involving expanding continental empires, such as the United States and Russia, might be biasing the results; yet the results hold if I confine the analysis to overseas colonies. I examine whether the distance between the metropole's capital and the location of a colonial war is associated with civilian harm. Yet neither the coefficient for distance, nor interaction terms that include it, achieve statistical significance. I also consider alternative model specifications, including models that consider individual explanatory variables on their own or that include different configurations of controls. In general, the coefficients for the core explanatory variables remain significant, with the sole exception of the combatant identity variable, which is sensitive to the inclusion of both the war aims and settler colony variables.

It is also possible that the findings might be driven by unmodeled temporal factors. Numerous scholars have noted the profound shifts in the strategic, economic, and normative context in which empire building took place from the nineteenth to twentieth centuries.Footnote 151 To explore this possibility, I add dichotomous variables for the post-1917 and the post-1945 periods, each of which has been posited as a turning point for practices of colonialism. Table 3 reports the results of these robustness checks. The coefficients for the post-1917 dummy variable (model 6) and the post-1945 dummy variable (model 7) do not achieve statistical significance, while the coefficients for the core explanatory variables remain unchanged. Collectively, these results suggest a surprising degree of continuity in the correlates of civilian victimization in colonial wars over time.

Table 3. Temporal Factors and Civilian Victimization in Colonial Wars

Notes: Robust standard errors in parentheses. *p < .05; **p < .01; ***p < .001.

While the correlates of civilian victimization in colonial wars appear consistent, it is possible that the strength of these associations has shifted over time. To explore this possibility, I sequentially add interaction terms for the post-1917 dummy and each of my core explanatory variables. While neither the combatant identity nor colonial institution interaction terms achieves statistical significance, the one for wartime strategy does (model 8). The predicted probability of civilian victimization is twelve points higher for guerrilla wars in the nineteenth century, but only five points higher in the twentieth century. There are two possible explanations for this decline. First, guerrilla tactics were the predominant way colonial wars were fought in the twentieth century, accounting for 84 percent of all colonial wars after 1917, which may decrease the importance of wartime strategy relative to other covariates. Second, twentieth-century colonial militaries refined their doctrines of “imperial policing,” which may have enhanced their ability to fight guerrilla wars without targeting civilians, although the brutality of many postwar counterinsurgency campaigns suggests that guerrilla tactics continued to pose profound challenges for metropolitan militaries.

Conclusion

Colonial powers claimed to be spreading civilization, yet frequently acted in barbaric ways in their colonial wars. In this article, I have argued that this penchant for brutality stems from the distinctive features of the colonial setting. Colonial powers often struggle to bring indigenous adversaries to battle, and are thus drawn to “scorched earth” methods. They frequently perceive their indigenous adversaries as racially inferior, and thus deserving targets of collective punishments. They govern imperial hierarchies through the fragile institutions of the colonial state, which can lead to insecurity, paranoia, and a tendency to respond to challenges with performative violence. Attempts to compensate for these liabilities, by recruiting settlers or indigenous intermediaries, can exacerbate these dynamics, providing additional pathways to civilian harm.

My analysis of 193 cases of colonial war between 1816 and 2003 confirms these hypotheses. Colonial powers are more likely than states in interstate wars to target civilians, to employ brutal methods, and to engage in mass killing. Variables related to the use of guerrilla methods, perceptions of racial difference, and the structure of the colonial state are all associated with civilian victimization in colonial wars across various model specifications and sensitivity analyses. Taken together, these findings modify the conventional wisdom regarding wartime civilian harm, which portrays it as a rational choice adopted for strategic reasons. In colonial wars, the normative and institutional setting is equally important. Colonial brutality is not just a wartime strategy, it is also a byproduct of racial hierarchies and imperial modes of governance.

These findings suggest a number of avenues for future research. First, they highlight the important role racial hierarchies can play in international politics. During the colonial era, race provided colonial powers with a framework for understanding why indigenous polities rebelled and what methods were required to suppress them. The experience of colonial warfare, in turn, hardened colonial understandings around race and elevated the maintenance of racial hierarchies into a central purpose of the colonial state. These racialized frameworks did not disappear with decolonization, however. They endure in the form of colonial-era ethnographies and manuals on “small wars,” which states draw on to guide their own contemporary counterinsurgency campaigns.Footnote 152 Exploring the linkages between colonial violence, racial understandings, evolving military doctrines, and varied practices of counterinsurgency is an important avenue of future research.

Second, the findings underscore the distinctiveness of the colonial state as a political unit in international politics. One of the primary imperatives of aspiring colonial state-builders was to manage violence to make up for their lack of legitimacy. The persistence of violent reactions to colonial state-building suggests that this project was necessarily contested and incomplete. The state structures that colonial powers handed over to their postcolonial successors were profoundly shaped by anxieties about—and often the experience of—colonial violence. Studies by Verghese and Mukherjee have illustrated how colonial rule left social and institutional legacies that shape patterns of political and ethnic violence in contemporary India.Footnote 153 Future studies could examine whether the experience of colonial violence, or different levels of colonial violence, has left similar historical legacies.

Finally, the findings shed light on why third-party interveners, who are often compared to colonial powers,Footnote 154 often struggle to fight nonstate adversaries in ways that avoid harming civilians. It is telling that in the three most recent “extra-state wars” in the COW data set—the 2000 Second Intifada, the 2001 Afghan Insurgency, and the 2003 Iraq Insurgency—the states involved have all been accused of mistreating civilians to varying degrees.Footnote 155 In all three cases, nonstate actors embraced guerrilla tactics and there were stark racial and religious divides between the primary antagonists. In the case of Israel, events in the West Bank highlight how the presence of armed settlers can fuel violent clashes. The American occupation of Iraq illustrates how the reliance on ethno-sectarian militias can unleash cycles of bloodletting. Reasonable people may disagree about whether these cases should be described as colonial wars, “internationalized” civil wars, or something else altogether. The bottom line is that the partitioning of data on political violence into three somewhat arbitrary categories, two of which are studied extensively and one of which is often overlooked, has limited our ability to draw meaningful comparisons. By ignoring colonial violence, we miss an opportunity to situate recent conflicts in their proper structural contexts and risk overlooking how racial prejudices and violent intermediaries can contribute to civilian suffering.

Data Availability Statement

Replication files for this article may be found at <https://doi.org/10.7910/DVN/9XV0B7>.

Supplementary Material

Supplementary material for this article is available at <https://doi.org/10.1017/S002081832300019X>.

Acknowledgments

Thank you to Bianca Freeman, Stacie Goddard, Robert Jervis, James McAllister, Michael MacDonald, Ngoni Munemo, Craig Murphy, Costantino Pischedda, Joseph Parent, Jack Snyder, Malik Izaak Taylor, Ari Weil, the editors, and two anonymous reviewers for their comments and discussions of the article's themes.

Footnotes

1. Adelman Reference Adelman2010, 408.

2. Dalrymple Reference Dalrymple2008, 7.

3. Groen Reference Groen2012, 289.

5. Etemad Reference Etemad2007, 93–94.

8. Valentino, Huth, and Balch-Lindsay Reference Valentino, Huth and Balch-Lindsay2004.

9. Downes Reference Downes2008, 35–36; Valentino Reference Valentino2004, chap. 6; Valentino, Huth, and Balch-Lindsay Reference Valentino, Huth and Balch-Lindsay2004, 393–94; Valentino, Huth, and Croco Reference Valentino, Huth and Croco2006, 370–72

10. Sarkees and Schafer Reference Sarkees and Schafer2000, 126.

11. Sarkees, Wayman, and Singer Reference Sarkees, Wayman and David Singer2003, 67.

12. Mansfield and Snyder Reference Mansfield and Snyder1995.

13. Fearon and Laitin Reference Fearon and Laitin2003.

14. Lyall and Wilson Reference Lyall and Wilson2009.

16. Arreguín-Toft Reference Arreguín-Toft2005.

17. A notable exception is Ravlo, Gleditsch, and Dorussen Reference Ravlo, Gleditsch and Dorussen2003, although their primary interest is in testing whether democratic states are more likely to participate in colonial wars. Katagiri Reference Katagiri2015 uses the COW extra-state data set to examine counterinsurgency outcomes, but does not center colonialism in his analysis. A number of studies have explored the relationship between violence and decolonization (see Goldsmith and He Reference Goldsmith and He2008; McAlexander Reference McAlexander2020; Paine Reference Paine2019), but they narrow their focus to the end of empire.

20. Ferguson Reference Ferguson2004; see also French Reference French2011, 133.

21. Boot Reference Boot2014, 124.

22. R.K. Edgerton Reference Edgerton2020, 167.

23. Thornton Reference Thornton2004, 88.

24. French Reference French2012, 751.

25. Daughton Reference Daughton2019, 522.

26. Hämäläinen Reference Hämäläinen2019, 368.

27. Creese Reference Creese2006, 28.

28. Quoted in Campbell Reference Campbell2018, 532.

29. Etemad Reference Etemad2007, 122.

31. Lawrence Reference Lawrence2010, 89–90.

33. Vandervort Reference Vandervort1998, 47–48.

35. MacDonald Reference MacDonald2014, 35–36.

37. French Reference French2012, 751.

38. Whittingham Reference Whittingham2012, 592.

39. McDougall Reference McDougall2005, 119.

40. Downes Reference Downes2008, 157–60.

41. Walter Reference Walter and Lewis2017, 36–37.

43. Lake and Reynolds Reference Lake and Reynolds2012.

44. Reid Reference Reid2007, 25.

45. Neep Reference Neep2012, 54.

46. Wagner Reference Wagner2018, 231.

49. Hobson and Sharman Reference Hobson and Sharman2005.

52. Conrad and Strange Reference Conrad, Strange and Risse2011, 42.

55. Burbank and Cooper Reference Burbank and Cooper2010, 325.

59. Müller-Crepon Reference Müller-Crepon2020, 710.

60. Moyd Reference Moyd2014, 141–43.

61. Mamdani Reference Mamdani1996, 37.

62. Lange Reference Lange2004, 906–907.

63. Lange Reference Lange2009, 29–30.

64. Buzan and Lawson Reference Buzan and Lawson2015, 131–32.

65. Sarkees, Wayman, and Singer Reference Sarkees, Wayman and David Singer2003, 58–59.

67. Arreguín-Toft Reference Arreguín-Toft2005; Lyall and Wilson Reference Lyall and Wilson2009.

69. Sarkees, Wayman, and Singer Reference Sarkees, Wayman and David Singer2003, 67–68.

70. Chamberlin Reference Chamberlin2018, chap. 1.

71. Downes Reference Downes2006, 156.

72. Morrow and Jo Reference Morrow and Jo2006.

73. This variable closely tracks civilian victimization, yet I retain it because it captures gradations in the relative intensity of civilian harm and facilitates comparison with Morrow and Jo's results.

74. Valentino Reference Valentino2004, 10–12.

75. Blacker Reference Blacker2007, 225.

77. The data for interstate wars are from Downes Reference Downes2008 and Morrow Reference Morrow2014.

78. Balcells and Kalyvas Reference Balcells and Kalyvas2014.

79. Salehyan, Siroky, and Wood Reference Salehyan, Siroky and Wood2014; Stanton Reference Stanton2016.

80. This difference is significant at the p < .001 level. All subsequent reported results are based on one-tailed t-tests for unpaired samples with unequal variance.

81. Downes Reference Downes2008, 25.

83. These differences are significant at the p < .001, p < .001, and p < .01 levels, respectively.

84. This difference lacks significance (p = .1612).

85. This difference is significant at the p < .01 level.

86. Valentino, Huth, and Balch-Lindsay Reference Valentino, Huth and Balch-Lindsay2004, 404.

87. Luttikhuis and Moses Reference Luttikhuis and Moses2012, 267.

88. Thomas Reference Thomas2005, 193.

89. While most colonial wars are fought by a single state, there are five cases of colonial powers fighting together, such as the 1816 Anglo–Dutch bombardment of Algiers.

90. Downes Reference Downes2008, 13–18.

95. DuBois Reference DuBois1925, 423.

98. Paine Reference Paine2019, 513–14.

99. Easterly and Levine Reference Easterly and Levine2016, 235.

100. Herbst Reference Herbst2000, 84–87; Lange Reference Lange2009, 47–49; Müller-Crepon Reference Müller-Crepon2020, 717–18.

101. Lange Reference Lange2009, 48.

102. Versions of these controls appear in Downes Reference Downes2008; Valentino, Huth, and Croco Reference Valentino, Huth and Croco2006; Valentino, Huth, and Balch-Lindsay Reference Valentino, Huth and Balch-Lindsay2004.

103. Porch Reference Porch and Paret1986, 394–95.

105. All predicted probabilities reported in these sections are based on the results of model 2, holding dichotomous variables at their modal values and continuous variables at their mean values.

106. Balcells and Kalyvas Reference Balcells and Kalyvas2014; Downes Reference Downes2008; Valentino, Huth, and Balch-Lindsay Reference Valentino, Huth and Balch-Lindsay2004.

107. Bidwell Reference Bidwell1973, 35.

108. Dhada Reference Dhada1998, 582.

109. Hack Reference Hack2022, 193–94.

110. Khodarkovsky Reference Khodarkovsky1992, 73.

111. Luong Reference Luong1992, 105.

112. Adelman Reference Adelman2010, 411.

113. Stanley Reference Stanley2022.

114. Board to Control to Government of India (13 February 1856), British Library IOR/E/4/834/801-852.

115. Ghosh Reference Ghosh1971, 88–89.

116. Thackwell Reference Thackwell1851, 212.

117. Baird Reference Baird1911, 57.

118. Fazal and Greene Reference Fazal and Greene2014.

119. Hämäläinen Reference Hämäläinen2019, 362.

122. Hull Reference Hull2005, 33.

123. Groen Reference Groen2012, 285.

124. Bennett Reference Bennett2007, 646–52.

125. Spiers Reference Spiers2004, 61.

126. Laband Reference Laband2014, 186.

127. Robbins Reference Robbins2012, 732.

128. Downes Reference Downes2008, chap. 5.

129. Pakenham Reference Pakenham1979, 536–47.

130. Abbenhuis Reference Abbenhuis2019, 104–111.

131. Miller Reference Miller2010, 319–20.

132. Thompson Reference Thompson2007, 112–13.

133. Marks Reference Marks1970, 185–86.

134. Guy Reference Guy2005, 103–104.

136. Webb Reference Webb2015, 47.

137. Jacoby Reference Jacoby2008, 128.

138. Thomas Reference Thomas2011, 144–45.

139. Loos Reference Loos1982, 61, 79.

140. Ndlovu-Gatsheni Reference Ndlovu-Gatsheni2009, 149.

141. Jung Reference Jung2007, 172.

142. Mockaitis Reference Mockaitis2012, 770.

143. Draper Reference Draper2019, 1027.

144. Brachet and Schelle Reference Brachet and Schelle2019, 61.

145. Adeleye Reference Adeleye1972, 207–208.

146. R.B. Edgerton Reference Edgerton1995, 228.

147. Linn Reference Linn1989, 145.

148. Hughes Reference Hughes2015, 594.

149. Lloyd Reference Lloyd2015, 306.

150. Raghaven Reference Raghaven2005, 254.

154. MacDonald Reference MacDonald2009.

155. Manekin Reference Manekin2020, chap. 7.

References

Abbenhuis, Maartje. 2019. The Hague Conferences and International Politics, 1898–1915. Bloomsbury.Google Scholar
Adams, Julia. 1996. Principals and Agents, Colonialists and Company Men: The Decay of Colonial Control in the Dutch East Indies. American Sociological Review 61 (1):1228.CrossRefGoogle Scholar
Adeleye, R.A. 1972. Mahdist Triumph and British Revenge in Northern Nigeria: Satiru 1906. Journal of the Historical Society of Nigeria 6 (2):193214.Google Scholar
Adelman, Jeremy. 2010. The Rites of Statehood: Violence and Sovereignty in Spanish America, 1789–1821. Hispanic American Historical Review 90 (3):391422.10.1215/00182168-2010-001CrossRefGoogle Scholar
Arreguín-Toft, Ivan. 2005. How the Weak Win Wars: A Theory of Asymmetric Conflict. Cambridge University Press.CrossRefGoogle Scholar
Baird, J.G.A. 1911. Private Letters of the Marquess of Dalhousie. Blackwood.Google Scholar
Balcells, Laia. 2017. Rivalry and Revenge: The Politics of Violence During Civil War. Cambridge University Press.10.1017/9781316392737CrossRefGoogle Scholar
Balcells, Laia, and Kalyvas, Stathis N.. 2014. Does Warfare Matter? Severity, Duration, and Outcomes of Civil Wars. Journal of Conflict Resolution 58 (8):1390–418.CrossRefGoogle Scholar
Balcells, Laia, and Stanton, Jessica A.. 2021. Violence Against Civilians During Armed Conflict: Moving Beyond the Macro- and Micro-Level Divide. Annual Review of Political Science 24:4569.CrossRefGoogle Scholar
Bell, Duncan. 2020. Dreamworlds of Race: Empire and the Utopian Destiny of Anglo-America. Princeton University Press.Google Scholar
Benjamin, Thomas. 2006. Encyclopedia of Western Colonialism Since 1450. Macmillan.Google Scholar
Bennett, Huw. 2007. The Other Side of the COIN: Minimum and Exemplary Force in British Army Counterinsurgency in Kenya. Small Wars and Insurgencies 18 (4):638–64.CrossRefGoogle Scholar
Bidwell, Robin. 1973. Morocco Under Colonial Rule: French Administration of Tribal Areas, 1912–1954. Frank Cass.Google Scholar
Blacker, John. 2007. The Demography of Mau Mau: Fertility and Mortality in Kenya in the 1950s. African Affairs 106 (423):205–27.10.1093/afraf/adm014CrossRefGoogle Scholar
Boot, Max. 2014. The Savage Wars of Peace: Small Wars and the Rise of American Power. Basic.Google Scholar
Brachet, Julien, and Schelle, Judith. 2019. The Value of Disorder: Autonomy, Prosperity, and Plunder in the Chadian Sahara. Cambridge University Press.CrossRefGoogle Scholar
Burbank, Jane, and Cooper, Frederick. 2010. Empires in World History: Power and the Politics of Difference. Princeton University Press.Google Scholar
Buzan, Barry, and Lawson, George. 2015. The Global Transformation: History, Modernity and the Making of International Relations. Cambridge University Press.CrossRefGoogle Scholar
Campbell, Caroline. 2018. Experiencing Colonial Violence from Below: French and Amazigh Entanglement During the Conquest of Morocco. French History 32 (4):532–53.CrossRefGoogle Scholar
Chamberlin, Paul. 2018. The Cold War's Killing Fields: Rethinking the Long Peace. Harper.Google Scholar
Clodfelter, Michael. 2008. Warfare and Armed Conflicts: A Statistical Encyclopedia of Casualty and Other Figures, 1494–2007. McFarland.Google Scholar
Conrad, Sebastian, and Strange, Marion. 2011. Governance and Colonial Rule. In Governance Without a State? Policies and Politics in Areas of Limited Statehood, edited by Risse, Thomas, 3964. Columbia University Press.Google Scholar
Crawford, Neta. 2002. Argument and Change in World Politics: Ethics, Decolonization, and Humanitarian Intervention. Cambridge University Press.CrossRefGoogle Scholar
Creese, Helen. 2006. A Puputan Tale: “The Story of the Pregnant Woman.” Indonesia 82:137.Google Scholar
Dalrymple, William. 2008. The Last Mughal: The Fall of a Dynasty, Delhi 1857. Vintage.Google Scholar
Daughton, J.P. 2019. The “Pacha Affair” Reconsidered: Violence and Colonial Rule in Interwar French Equatorial Africa. Journal of Modern History 91 (3):493524.CrossRefGoogle Scholar
Dhada, Mustafah. 1998. The Liberation War in Guinea-Bissau Reconsidered. Journal of Military History 62 (3):517–93.CrossRefGoogle Scholar
Downes, Alexander B. 2006. Desperate Times, Desperate Measures: The Causes of Civilian Victimization in War. International Security 30 (4):152–95.CrossRefGoogle Scholar
Downes, Alexander B. 2008. Targeting Civilians in War. Cornell University Press.Google Scholar
Draper, Mario. 2019. The Force Publique's Campaigns in the Congo–Arab War, 1892, 1894. Small Wars and Insurgencies 30 (4/5):1020–39.CrossRefGoogle Scholar
DuBois, W.E.B. 1925. Worlds of Color. Foreign Affairs 3 (3):423–44.CrossRefGoogle Scholar
Easterly, William, and Levine, Ross. 2016. The European Origins of Economic Development. Journal of Economic Growth 21:225–57.CrossRefGoogle Scholar
Edgerton, Robert B. 1995. The Fall of the Asante Empire. Free Press.Google Scholar
Edgerton, Ronald K. 2020. American Datu: John J. Pershing and Counterinsurgency Warfare in the Muslim Philippines, 1899–1913. University Press of Kentucky.Google Scholar
Etemad, Bouda. 2007. Possessing the World: Taking the Measurements of Colonisation from the Eighteenth to the Twentieth Century. Berghahn Books.Google Scholar
Fanon, Frantz. 2004. The Wretched of the Earth. Translated by Philcox, Richard. Grove Press.Google Scholar
Fazal, Tanisha M. 2018. Wars of Law: Unintended Consequences in the Regulation of Armed Conflict. Cornell University Press.Google Scholar
Fazal, Tanisha M., and Greene, Brooke C.. 2014. A Particular Difference: European Identity and Civilian Targeting. British Journal of Political Science 45 (4):829–51.CrossRefGoogle Scholar
Fearon, James D., and Laitin, David D.. 2003. Ethnicity, Insurgency, and Civil War. American Political Science Review 97 (1):7590.CrossRefGoogle Scholar
Ferguson, Niall. 2004. Empire: The Rise and Demise of the British World Order and the Lessons for Global Power. Basic Books.Google Scholar
Ferris, John. 2012. Small Wars and Great Games: The British Empire and Hybrid Warfare, 1700–1970. In Hybrid Warfare: Fighting Complex Opponents from the Ancient World to the Present, edited by Murray, Williamson and Mansoor, Peter R., 199224. Cambridge University Press.CrossRefGoogle Scholar
Freeman, Bianca. 2023. Racial Hierarchy and Jurisdiction in US Status of Forces Agreements. Working paper, University of California, San Diego.Google Scholar
French, David. 2011. The British Way in Counter-Insurgency, 1945–1967. Oxford University Press.CrossRefGoogle Scholar
French, David. 2012. Nasty Not Nice: British Counter-Insurgency Doctrine and Practice, 1945–1967. Small Wars and Insurgencies 23 (4/5):774–61.10.1080/09592318.2012.709763CrossRefGoogle Scholar
Ghosh, Suresh Chandra. 1971. Dalhousie and the Santal Insurrection of 1855. Bengal Past and Present 40 (169):8598.Google Scholar
Goldsmith, Benjamin E., and He, Baogang. 2008. Letting Go Without a Fight: Decolonization, Democracy and War, 1900–1994. Journal of Peace Research 45 (5):587611.CrossRefGoogle Scholar
Green, Amelia Hoover. 2018. The Commander's Dilemma: Violence and Restraint in Wartime. Cornell University Press.10.7591/cornell/9781501726477.001.0001CrossRefGoogle Scholar
Groen, Petra. 2012. Colonial Warfare and Military Ethics in the Netherlands East Indies, 1816–1941. Journal of Genocide Research 14 (3/4):277–96.CrossRefGoogle Scholar
Guha, Ranajit. 1988. The Prose of Counterinsurgency. In Selected Subaltern Studies, edited by Guha, Ranajit and Spivak, Gayatri Chakravorty, 4588. Oxford University Press.Google Scholar
Guy, Jeff. 2005. The Maphumulo Uprising: War, Law and Ritual in the Zulu Rebellion. University of KwaZulu-Natal Press.Google Scholar
Hack, Karl. 2022. The Malayan Emergency: Revolution and Counterinsurgency at the End of Empire. Cambridge University Press.Google Scholar
Hämäläinen, Pekka. 2019. Lakota America: A New History of Indigenous Power. Yale University Press.10.2307/j.ctvqc6gp2CrossRefGoogle Scholar
Herbst, Jeffrey. 2000. States and Power in Africa: Comparative Lessons in Authority and Control. Princeton University Press.Google Scholar
Hobson, John M., and Sharman, J.C.. 2005. The Enduring Place of Hierarchy in World Politics: Tracing the Social Logics of Hierarchy and Political Change. European Journal of International Relations 11 (1):6398.CrossRefGoogle Scholar
Hughes, Matthew. 2009. The Banality of Brutality: British Armed Forces and the Repression of the Arab Revolt in Palestine, 1936–39. English Historical Review 507:313–54.CrossRefGoogle Scholar
Hughes, Matthew. 2015. Terror in Galilee: British–Jewish Collaboration and the Special Night Squads in Palestine During the Arab Revolt, 1938–39. Journal of Imperial and Commonwealth History 43 (4):590610.CrossRefGoogle Scholar
Hull, Isabel V. 2005. Absolute Destruction: Military Culture and the Practices of War in Imperial Germany. Cornell University Press.Google Scholar
Humphreys, Macartan, and Weinstein, Jeremy M.. 2006. Handling and Manhandling Civilians in Civil War. American Political Science Review 100 (3):429–47.CrossRefGoogle Scholar
Jacoby, Karl. 2008. Shadows at Dawn: An Apache Massacre and the Violence of History. Penguin.Google Scholar
Jung, Patrick J. 2007. The Black Hawk War of 1832. University of Oklahoma Press.Google Scholar
Kalyvas, Stathis N. 2006. The Logic of Violence in Civil War. Cambridge University Press.CrossRefGoogle Scholar
Kanya-Forstner, A.S. 1989. The French Marines and the Conquest of the Western Sudan, 1880–1899. In Imperialism and War: Essays on Colonial Wars in Asia and Africa, edited by De Moor, J.A. and Wesseling, H.L., 121–45. Leiden: Brill.CrossRefGoogle Scholar
Katagiri, Noriyuki. 2015. Adapting to Win: How Insurgents Fight and Defeat Foreign States in War. University of Pennsylvania Press.CrossRefGoogle Scholar
Khodarkovsky, Michael. 1992. Bitter Choices: Loyalty and Betrayal in the Russian Conquest of the North Caucasus. Cornell University Press.Google Scholar
Kinsella, Helen M. 2011. The Image Before the Weapon: A Critical History of the Distinction Between Combatant and Civilian. Cornell University Press.CrossRefGoogle Scholar
Kinsella, Helen M. 2023. Settler Empire and the United States: Francis Lieber and the Laws of War. American Political Science Review 117 (2):629–42.CrossRefGoogle Scholar
Kolsky, Elizabeth. 2015. The Colonial Rule of Law and the Legal Regime of Exception: Frontier “Fanaticism” and State Violence in British India. American Historical Review 120 (4):1218–46.CrossRefGoogle Scholar
Laband, John. 2014. The Transvaal Rebellion: The First Boer War, 1880–1881. Routledge.CrossRefGoogle Scholar
Lake, Marilyn, and Reynolds, Henry. 2012. Drawing the Global Colour Line: White Men's Countries and the Question of Racial Equality. Cambridge University Press.Google Scholar
Lange, Matthew K. 2004. British Colonial Legacies and Political Development. World Development 32 (6):905–22.CrossRefGoogle Scholar
Lange, Matthew K. 2009. Lineages of Despotism and Development: British Colonialism and State Power. Cambridge University Press.CrossRefGoogle Scholar
Lawrence, Adria. 2010. Triggering Nationalist Resistance: Competition and Conflict in Uprisings Against Colonial Rule. International Security 35 (2):88122.CrossRefGoogle Scholar
Linn, Brian McAllister. 1989. The US Army and Counterinsurgency in the Philippine War, 1899–1902. University of North Carolina Press.Google Scholar
Lloyd, Nick. 2015. Colonial Counter-Insurgency in Southern India: The Malabar Rebellion, 1921–1922. Contemporary British History 29 (3):297317.CrossRefGoogle Scholar
Loos, Noel. 1982. Invasion and Resistance: Aboriginal–European Relations on the North Queensland Frontier, 1861–1897. Australian National University Press.Google Scholar
Luong, Hy V. 1992. Revolution in the Village: Tradition and Transformation in North Vietnam, 1925–1928. University of Hawaii Press.CrossRefGoogle Scholar
Luttikhuis, Bart, and Moses, A. Dirk. 2012. Mass Violence and the End of the Dutch Colonial Empire in Indonesia. Journal of Genocide Research 14 (3/4):256–76.CrossRefGoogle Scholar
Lyall, Jason, and Wilson, Isaiah. 2009. Rage Against the Machines: Explaining Outcomes in Counterinsurgency Wars. International Organization 63 (1):67106.CrossRefGoogle Scholar
MacDonald, Paul K. 2009. Those Who Forget Historiography Are Doomed to Republish It: Empire, Imperialism and Contemporary Debates About American Power. Review of International Studies 35 (1):4567.CrossRefGoogle Scholar
MacDonald, Paul K. 2013. “Retribution Must Succeed Rebellion”: The Colonial Origins of Counterinsurgency Failure. International Organization 67 (2):253–86.CrossRefGoogle Scholar
MacDonald, Paul K. 2014. Networks of Domination: The Social Foundations of Peripheral Conquest in International Politics. Oxford University Press.CrossRefGoogle Scholar
MacDonald, Paul K. 2018. Embedded Authority: A Relational Network Approach to Hierarchy in World Politics. Review of International Studies 44 (1):128–50.10.1017/S0260210517000213CrossRefGoogle Scholar
MacDonald, Paul K. 2020. The Governor's Dilemma in Colonial Empires. In The Governor's Dilemma: Indirect Governance Beyond Principals and Agents, edited by Abbott, Kenneth W., Genschel, Philipp, Snidal, Duncan, and Zangl, Bernhard, 3958. Oxford University Press.CrossRefGoogle Scholar
Mamdani, Mahmood. 1996. Citizen and Subject: Contemporary Africa and the Legacy of Late Colonialism. Princeton University Press.Google Scholar
Mamdani, Mahmood. 2001. When Victims Become Killers: Colonialism, Nativism, and the Genocide in Rwanda. Princeton University Press.Google Scholar
Manchanda, Nivi. 2020. Imagining Afghanistan: The History and Politics of Imperial Knowledge. Cambridge University Press.CrossRefGoogle Scholar
Manekin, Devorah S. 2020. Regular Soldiers, Irregular War: Violence and Restraint in the Second Intifada. Cornell University Press.Google Scholar
Mansfield, Edward D., and Snyder, Jack L.. 1995. Democratization and the Danger of War. International Security 20 (1):538.CrossRefGoogle Scholar
Marks, Shula. 1970. Reluctant Rebellion: The 1906–1908 Disturbances in Natal. Oxford University Press.Google Scholar
McAlexander, Richard J. 2020. The Politics of Anticolonial Resistance: Violence, Nonviolence, and the Erosion of Empire. PhD diss., Columbia University.10.33774/apsa-2020-m93gcCrossRefGoogle Scholar
McDougall, James. 2005. Savage Wars? Codes of Violence in Algeria, 1830s–1990s. Third World Quarterly 26 (1):117–31.CrossRefGoogle Scholar
Metcalf, Thomas R. 1995. Ideologies of the Raj. Cambridge University Press.CrossRefGoogle Scholar
Miller, Stephen M. 2010. Duty or Crime? Defining Acceptable Behavior in the British Army in South Africa, 1899–1902. Journal of British Studies 49 (2):311–31.CrossRefGoogle Scholar
Miller, Stephen M. 2013. The British Way of War: Cultural Assumptions and Practices in the South African War, 1899–1902. Journal of Military History 77 (4):1329–47.Google Scholar
Mockaitis, Thomas R. 2012. The Minimum Force Debate: Contemporary Sensibilities Meet Imperial Practice. Small Wars and Insurgencies 23 (4/5):762–80.CrossRefGoogle Scholar
Morrow, James D. 2007. When Do States Follow the Laws of War? American Political Science Review 101 (3):559–72.CrossRefGoogle Scholar
Morrow, James D. 2014. Order Within Anarchy: The Laws of War As an International Institution. Cambridge University Press.CrossRefGoogle Scholar
Morrow, James D., and Jo, Hyeran. 2006. Compliance with the Laws of War: Dataset and Coding Rules. Conflict Management and Peace Science 23 (1):91113.CrossRefGoogle Scholar
Moyd, Michelle R. 2014. Violent Intermediaries: African Soldiers, Conquest, and Everyday Colonialism in German East Africa. Ohio University Press.Google Scholar
Mukherjee, Shivaji. 2021. Colonial Institutions and Civil War: Indirect Rule and Maoist Insurgency in India. Cambridge University Press.CrossRefGoogle Scholar
Müller-Crepon, Carl. 2020. Continuity or Change? (In)direct Rule in British and French Colonial Africa. International Organization 74 (4):707–41.CrossRefGoogle Scholar
Ndlovu-Gatsheni, Sabelo. 2009. The Ndebele Nation: Reflections on Hegemony, Memory, and Historiography. Rozenberg.Google Scholar
Neep, Daniel. 2012. Occupying Syria Under the French Mandate: Insurgency, Space, and State Formation. Cambridge University Press.CrossRefGoogle Scholar
Paine, Jack. 2019. Redistributive Political Transitions: Minority Rule and Liberation Wars in Colonial Africa. Journal of Politics 81 (2):505–23.CrossRefGoogle Scholar
Pakenham, Thomas. 1979. The Boer War. Random House.Google Scholar
Porch, Douglas. 1986. Bugeaud, Galliéni, Lyautey: The Development of French Colonial Warfare. In Makers of Modern Strategy from Machiavelli to the Nuclear Age, edited by Paret, Peter, 376407. Princeton University Press.CrossRefGoogle Scholar
Provence, Michael. 2015. French Mandate Counterinsurgency and the Repression of the Great Syrian Revolt. In Routledge Handbook of the History of the Middle East Mandates, edited by Schayegh, Cyrus and Arsan, Andrew, 136–52. Routledge.Google Scholar
Raghaven, Srinath. 2005. Protecting the Raj: The Army in India and Internal Security, c. 1919–39. Small Wars and Insurgencies 16 (3):253–79.CrossRefGoogle Scholar
Rand, Gavin. 2017. “From the Black Mountain to Waziristan”: Culture and Combat on the North-West Frontier. In Culture, Conflict, and the Military in Colonial South Asia, edited by Roy, Kaushik and Rand, Gavin, 131–56. Routledge.CrossRefGoogle Scholar
Ravlo, Hilde, Gleditsch, Nils Petter, and Dorussen, Han. 2003. Colonial War and Democratic Peace. Journal of Conflict Resolution 47 (4):520–48.10.1177/0022002703254295CrossRefGoogle Scholar
Reid, Richard J. 2007. Revisiting Primitive War: Perceptions of Violence and Race in History. War and Society 26 (2):125.CrossRefGoogle Scholar
Reid, Richard J. 2012. Warfare in African History. Cambridge University Press.CrossRefGoogle Scholar
Rid, Thomas. 2010. The Nineteenth Century Origins of Counterinsurgency Doctrine. Journal of Strategic Studies 33 (5):727–58.CrossRefGoogle Scholar
Robbins, Simon. 2012. The British Counter-Insurgency in Cyprus. Small Wars and Insurgencies 23 (4/5):720–43.CrossRefGoogle Scholar
Robinson, Ronald E. 1972. Non-European Foundations of European Imperialism. In Studies in the Theory of Imperialism, edited by Owen, Roger and Sutcliffe, Robert, 117–42. Longman.Google Scholar
Salehyan, Idean, Siroky, David, and Wood, Reed M.. 2014. External Rebel Sponsorship and Civilian Abuse: A Principal-Agent Analysis of Wartime Atrocities. International Organization 68 (3):633–61.CrossRefGoogle Scholar
Sarkees, Meredith Reid, and Schafer, Phil. 2000. The Correlates of War Data on War: An Update to 1997. Conflict Management and Peace Science 18 (1):123–44.CrossRefGoogle Scholar
Sarkees, Meredith Reid, Wayman, Frank Whelon, and David Singer, J.. 2003. Inter-State, Intra-State, and Extra-State Wars: A Comprehensive Look at Their Distribution over Time, 1816–1997. International Studies Quarterly 47 (1):4970.CrossRefGoogle Scholar
Sharman, J.C. 2020. Empires of the Weak: The Real Story of European Expansion and the Creation of the New World Order. Princeton University Press.Google Scholar
Spiers, Edward M. 2004. The Victorian Soldier in Africa. Manchester University Press.CrossRefGoogle Scholar
Spruyt, Hendrick. 2005. Ending Empire: Contested Sovereignty and Territorial Partition. Cornell University Press.CrossRefGoogle Scholar
Stanley, Peter. 2022. Hul! Hul! The Suppression of the Santal Rebellion in Bengal, 1855. Hurst.Google Scholar
Stanton, Jessica A. 2016. Violence and Restraint in Civil War: Civilian Targeting in the Shadow of International Law. Cambridge University Press.CrossRefGoogle Scholar
Thackwell, Edward Joseph. 1851. Narrative of the Second Seikh War in 1848–49. Richard Bentley.Google Scholar
Thomas, Martin. 2005. The French Empire Between the Wars: Imperialism, Politics, and Society. Manchester University Press.Google Scholar
Thomas, Martin. 2011. Colonial Violence in Algeria and the Distorted Logic of State Retribution: The Sétif Uprising of 1945. Journal of Military History 75 (1):125–57.Google Scholar
Thompson, Paul S. 2007. Crossroads of War: The People of Nkandla in the Zulu Rebellion of 1906. Scientia Militaria 35 (2):95127.Google Scholar
Thornton, Rod. 2004. The British Army and the Origins of Its Minimum Force Philosophy. Small Wars and Insurgencies 15 (1):83106.CrossRefGoogle Scholar
Valentino, Benjamin A. 2004. Final Solutions: Mass Killing and Genocide in the Twentieth Century. Cornell University Press.Google Scholar
Valentino, Benjamin A. 2014. Why We Kill: The Political Science of Political Violence Against Civilians. Annual Review of Political Science 17:89103.CrossRefGoogle Scholar
Valentino, Benjamin, Huth, Paul, and Balch-Lindsay, Dylan. 2004. “Draining the Sea”: Mass Killing and Guerrilla Warfare. International Organization 58 (2):375407.CrossRefGoogle Scholar
Valentino, Benjamin, Huth, Paul, and Croco, Sarah. 2006. Covenants Without the Sword: International Law and the Protection of Civilians in Times of War. World Politics 58 (3):339–77.CrossRefGoogle Scholar
Vandervort, Bruce. 1998. Wars of Imperial Conquest in Africa, 1830–1914. Indiana University Press.Google Scholar
Verghese, Ajay. 2016. The Colonial Origins of Ethnic Violence in India. Stanford University Press.Google Scholar
Vogt, Manuel, Bormann, Nils-Christian, Rüegger, Seraina, Cederman, Lars-Erik, Hunziker, Philipp, and Girardin, Luc. 2015. Integrating Data on Ethnicity, Geography, and Conflict: The Ethnic Power Relations Data Set. Journal of Conflict Resolution 59 (7):1327–42.CrossRefGoogle Scholar
Wagner, Kim A. 2018. Savage Warfare: Violence and the Rule of Colonial Difference in Early British Counterinsurgency. History Workshop Journal 85 (1):217–37.CrossRefGoogle Scholar
Walter, Dierk. 2017. Colonial Violence: European Empires and the Use of Force. Translated by Lewis, Peter. Oxford University Press.Google Scholar
Webb, Denver A. 2015. War, Racism, and the Taking of Heads: Revisiting Military Conflict in the Cape Colony and Western Xhosaland in the Nineteenth Century. Journal of African History 56 (1):3755.CrossRefGoogle Scholar
Weinstein, Jeremy M. 2007. Inside Rebellion: The Politics of Insurgent Violence. Cambridge University Press.Google Scholar
Whittingham, Daniel. 2012. “Savage Warfare”: C.E. Callwell, the Roots of Counter-Insurgency, and the Nineteenth Century Context. Small Wars and Insurgenices 23 (4/5):591607.CrossRefGoogle Scholar
Wolfe, Patrick. 2006. Settler Colonialism and the Elimination of the Native. Journal of Genocide Research 8 (4):387409.CrossRefGoogle Scholar
Figure 0

Figure 1. Frequency of Colonial and Interstate Wars, 1816–2003

Figure 1

Table 1. Comparing Violence in Colonial Versus Interstate Wars, 1816–2003

Figure 2

Table 2. Correlates of Civilian Victimization in Colonial Wars

Figure 3

Figure 2. Marginal Effects of Core Explanatory Variables

Figure 4

Table 3. Temporal Factors and Civilian Victimization in Colonial Wars

Supplementary material: File

MacDonald supplementary material
Download undefined(File)
File 820.3 KB
Supplementary material: File

MacDonald_Dataset

Dataset

Download MacDonald_Dataset(File)
File