View source for Human Sacrifice

<!-- you can have any number of categories here -->
[[Category:Peter Leeson]]
[[Category:Make Or Break Views Of Libertarianism]]
[[Category:Austrian Economics]]
[[Category:Rational Choice Theory]]
[[Category:GIGO]]
<!-- 1 URL must be followed by >= 0 Other URL and Old URL and 1 End URL.-->
{{URL | url = http://www.peterleeson.com/Human_Sacrifice.pdf}}
<!-- {{Other URL | url = }} -->
<!-- {{Old URL | url = }} -->
{{End URL}}
{{DES | des = [[Peter Leeson]] argues that "that human sacrifice is a technology for protecting property rights."  A wild-ass theory that ignores other possible social explanations, relying on pre-anthropological stories.  Amusingly, this article claims the voluntary acceptance of British dominance as a means of ending local wars in exchange for ending human sacrifice. | show=}}
<!-- insert wiki page text here -->
<!-- DPL has problems with categories that have a single quote in them.  Use these explicit workarounds. -->
<!-- otherwise, we would use {{Links}} and {{Quotes}}  -->
{{List|title=Human Sacrifice|links=true}}
{{Quotations|title=Human Sacrifice|quotes=true}}

{{Text |
ABSTRACT
This paper develops a theory of rational human sacrifice: the purchase and ritual slaughter of innocent persons to appease divinities. I argue that human sacrifice is a technology for protecting property rights. It improves prop- erty protection by destroying part of sacrificing communities’ wealth, which depresses the expected payoff of plundering them. Human sacrifice is a highly effective vehicle for destroying wealth to protect property rights because it is an excellent public meter of wealth destruction. Human sacrifice is spectacu- lar, publicly communicating a sacrificer’s destruction far and wide. Further, immolating a live person is nearly impossible to fake, verifying the amount of wealth a sacrificer has destroyed. To incentivize community members to contribute wealth for destruction, human sacrifice is presented as a religious obligation. To test my theory I investigate human sacrifice as practiced by the most significant and well-known society of ritual immolators in the modern era: the Konds of Orissa, India. Evidence from the Konds supports my theory’s predictions.
∗ I thank Doug Allen, Yoram Barzel, Pete Boettke, Chris Coyne, Alex Fink, Jayme Lemke, Barkley Rosser, Jesse Shapiro, and David Skarbek for helpful comments and conversation.
ISSN 2326-6198; DOI 10.1561/105.00000007 ⃝c 2014 P. T. Leeson

138
Leeson
“[E]very durable social institution or practice is efficient.”
— George J. Stigler (1992, p. 459)
“[W]here we find a warlike ferocious race, delighting in cruelty and dev- astation, we may be assured that they will have deities delighting in slaughter, and rites polluted with blood.”
— C.R. (1846a, p. 60)
1 Introduction
Few practices evoke the shock and horror of human sacrifice: the purchase and ritual slaughter of innocent persons to appease divinities. And few practices seem more obviously irrational. Indeed, purchasing persons only to murder them to satisfy the blood thirst of imaginary beings might be considered the apotheosis of irrationality.1
Even the economists who have mentioned human sacrifice take this view. One asserts that human sacrifice is “properly considered as noneconomic” and thus beyond the power of rational choice theory to illuminate (Hunter et al., 1957, p. 59). A more recent reference adduces human sacrifice in support of behavioralist doubts about the canonical rendition of “economic man” (Ainslie, 2005, p. 816).
My paper takes the opposite view. It develops a theory of rational human sacrifice, tests that theory, and explains the practice of human sacrifice among the same people whose reliance on ritual immolation the work pointed to above suggests is inexplicable using orthodox economics.
I argue that human sacrifice is a technology for protecting property rights. My theory builds on Allen (2002) who introduced the idea that lowering an asset’s gross value might sometimes be a sensible way to improve its enforcement. My theory is simple.
In agricultural societies nature produces variation in land’s output. This variation creates disparities between communities’ wealth. Absent government, wealth dis- parities induce conflict between communities, as those occupying land that received a relatively unfavorable natural shock seek to plunder those whose expected wealth is higher. If conflict’s cost is sufficiently high, it is cheaper for communities to pro- tect their property rights by destroying part of their wealth. Wealth destruction
1 In contrast, human sacrifice as, for instance, the Aztecs famously practiced it is neither sur- prising nor mysterious. The Aztec’s sacrificial victims were overwhelming one of two sorts: captured enemy soldiers and criminals. Here human sacrifice was little more than capital pun- ishment. This article exclusively considers the puzzling practice of immolating innocent persons purchased only for that purpose.

Human Sacrifice 139
depresses the expected payoff of plunder and in doing so protects rights in wealth that remains.
Human sacrifice is a method of such destruction. By exchanging valuable prop- erty for humans and then slaughtering them, communities destroy wealth. Human sacrifice is a highly effective vehicle for destroying wealth to protect property rights because it is an excellent public meter of wealth destruction.
Unlike burning a mound of crops, human sacrifice is spectacular, communicating a sacrificing community’s wealth destruction far and wide. This permits would-be predators who do not observe destruction directly to learn of it nonetheless. Further, unlike burning a mound of crops, which can be manipulated to appear to destroy more wealth than is in fact destroyed, immolating a live human is nearly impossible to fake. This ensures would-be predators that seemingly destroyed wealth is in fact destroyed.
To incentivize community members to contribute wealth for destruction, human sacrifice is presented as a religious obligation. Persons who believe that they must contribute valuable property toward the purchase of sacrificial victims to procure divinities’ good favor are more likely to do so.
The logic my theory develops is not limited to understanding the practice of human sacrifice. It illuminates a wide variety of behavior seemingly at odds with the economic maxim “more is preferred to less” — from the mundane to the extraordi- nary. For example, it helps explain why some persons who live in unsafe neighbor- hoods drive cheaper cars than persons with similar incomes who live in safer ones; why many businesses in developing economies deliberately eschew growth; why per- sons in ancient societies expended inordinate resources building monumental tombs, such as pyramids; and why these same persons buried their most valuable goods with the dead. Each of these behaviors is marked by what might be called “conspic- uous destruction”: persons deliberating choosing to make themselves poorer than they could otherwise be.
Such destruction is especially striking where we find groups of persons pursu- ing outright poverty. For example, Scott (2009) documents this pursuit among the Zomia people of Southeast Asia who choose to live at near subsistence levels. Gyp- sies, who have traditionally prevented themselves from accumulating greater wealth by refusing even basic literacy, might exemplify such a group too.2 So may the members of ascetic groups, such as monks, who take vows of poverty, or the Jewish Essenes, who, famously, lived in voluntary destitution for centuries.
2 On the economics of Gypsy practices, see Leeson (2013).

140 Leeson
The logic my theory of human sacrifice develops might also be used to illuminate the impoverishing choices of these groups’ members. It suggests that what distin- guishes them from others may not be different preferences for material wealth, but rather different vulnerability to predation that renders conspicuous destruction a rational means of protecting what little they do have from the grasping hands of others.
To investigate my theory I consider human sacrifice as practiced by the most significant and well-known society of ritual immolators in the modern era: the Konds of Orissa, India. The western world stumbled upon the Konds in 1835 when the British government sought to capture the raja of Goomsur, a zemindary of Orissa, who refused to pay revenues he owed the British East India Company. The raja fled from his home in the plains of Orissa to the remote hills above the Eastern Ghats where the British followed him.
Forced to tread into these hills for the first time, the British “discovered, . . . with mingled horror and surprise, that a system of human sacrifice, aggravated by the cruel manner of its performance, existed almost universally” (Campbell, 1861, p. 51; see also, Dalton, 1872, p. 285).3 After successfully prosecuting several wars in Kond country in search of the raja, the British turned their attention to eradicating the “atrocious rite of sacrificing human victims” among the Konds (Campbell, 1861, p. 47). This effort culminated in the establishment of an official British agency for the suppression of human sacrifice in the hill tracts of Orissa.
At different times two British officers led this agency’s operations: Major Samuel Macpherson and Major John Campbell. Both agents left government reports, fur- nished those reports’ information to periodicals, and later themselves published works describing the Konds, the Konds’ institution of human sacrifice, and the British agency’s efforts to stamp out this institution among them.
My discussion is based on these men’s invaluable writings, which constitute the most important primary source documents relating to human sacrifice among the Konds. To supplement these records I draw on primary source documents from other first-hand observers of the Konds and numerous secondary sources dealing with the Konds and their institution of human sacrifice.
The Konds permit me to test several predictions of my theory. Evidence from their institution of human sacrifice supports these predictions.
3 The British were also horrified to learn that some Kond communities practiced female infan- ticide. They sought to suppress this custom where practiced, together with human sacrifice.

Human Sacrifice 141
My analysis contributes to the literature that, following Demsetz’s (1967) sem- inal work, explores the private emergence and enforcement of property rights.4 Johnsen’s (1986) contribution to this literature, which analyzes the potlatch system of wealth redistribution, is of special relevance given the important role that the observability of what one does with his wealth plays in my theory. Johnsen describes how, through potlatch, Southern Kwakiutl Native Americans prone to conflict pub- licly transferred wealth among themselves, improving their property protection.5 My article explains how property rights can be and have been protected privately through a more unsuspecting source: human sacrifice.
2 A Theory of Human Sacrifice
2.1 Property Protection through Violent Conflict
Consider an agricultural society populated by an arbitrarily large, even number of communities. This society has no government. Its communities interact an indefinite number of periods.
Each community consists of the same number of identical, risk-neutral persons and occupies an identical tract of land. The communities value a tract of land at t > 0. t is the same every period.
Each tract of land generates physically identical output, subject to random, nature-produced shocks, such as weather and animal activity. These shocks create random variation in the quantity of output a community enjoys each period.
Two sorts of such shocks are possible: a relatively unfavorable shock, which generates a relatively low quantity of output, and a relatively favorable shock, which generates a relatively high quantity of output. Every period nature shocks each community relatively favorably or unfavorably with equal probability. I call
4 See, for instance, Friedman (1979), Benson (1989), Anderson and Hill (2004), Leeson (2007), and Skarbek (2011).
5 Potlatch is useful for this purpose when groups’ wealth is common knowledge and when poorer groups cannot threaten richer ones. When groups’ wealth is private information, as it is in the case I consider, all groups have an incentive to refuse to make transfers on the grounds that they are poor, and thus should rather be transfer recipients. Further, when threats are possi- ble, poorer groups can use the transfer system to extort richer ones, encouraging rather than discouraging property aggression. Under these circumstances, wealth destruction, in contrast, is effective. Notably, Johnsen (1986) indicates that American Indian potlatch sometimes took the form of wealth destruction rather than wealth redistribution. Though he analyzes only the former.

142 Leeson
a community that receives a relatively favorable natural shock in a period a lucky community and a community that receives a relatively unfavorable natural shock in a period an unlucky community.
Communities value a given quantity of output equally. gh > 0 is the value of a lucky community’s output. gl > 0 is the value of an unlucky community’s output where gh = 3gl + 2t.
g and t are in the same units. Together they comprise a community’s wealth in a period, vh = t+gh and vl = t+gl, where vh = 3vl. Each community knows how nature shocked it in a period but not how nature shocked other communities the same period. Thus communities privately observe their wealth.
After assigning communities wealth, each period nature randomly assigns each community a “neighbor” — another community it interacts with. A community interacts only with its neighbor each period. Each period has three potential stages:
1. Nature randomly assigns each community wealth.
2. Nature randomly assigns each community a neighbor and each community
chooses how to interact with its neighbor.
A community has two choices: it may attempt to seize its neighbor’s wealth through extortion, demanding a price from its neighbor under the threat of violent attack if its neighbor refuses to pay, or it may not attempt to seize its neighbor’s wealth, in which case it does nothing. Communities make their choices simultane- ously. If both neighbors do nothing, each earns its nature-assigned wealth. If at least one neighbor attempts to extort the other, there is a third stage where each threatened community has two subsequent choices.
3. A threatened community chooses whether to surrender to its neighbor, paying the price its neighbor demands, or to refuse to pay its neighbor the price it demands. Threatened communities make their choices simultaneously.
If at least one threatened community in a pair of neighbors surrenders, there is peace. In this case a surrendering community earns its nature-assigned wealth less the extortion price its neighbor demands. A surrendering community’s neighbor earns its nature-assigned wealth plus the extortion price its neighbor pays.
If no threatened community in a pair of neighbors surrenders, there is a war. In this case both communities earn their expected payoff from violent conflict with their neighbor. War is costly. It costs each warring neighbor c = 9vl/20, which may be thought of as the cost associated with community members dying in conflict.
The only means of warfare in this society is hand-to-hand combat. Since each community has the same number of identical persons, if neighbors war, each wins
Human Sacrifice 143
the violent conflict with the same probability: 1/2. A war’s winner earns its own wealth plus the loser’s minus the cost of war. A war’s loser earns zero minus the cost of war.
Communities completely consume their end-period wealth when their interactions are concluded before the next period begins. Between periods nature repopulates communities whose members died in conflict and assigns communities that lost wars new tracts of land such that every community begins every period in the same position it is in the first period before nature randomly assigns wealth. All the information described above is commonly known.
Using this information it is easy to tabulate each community’s expected payoff of war. An unlucky community earns:
1􏰀v +v􏰁 1􏰀v+v􏰁 21v
h l + l l −c= l. (1)
2 2 2 2 20 A lucky community earns:
1􏰀v +v 􏰁 1􏰀v +v 􏰁 41v
l h + h h −c= l. (2)
The minimum price that an unlucky community would demand from its neighbor if it threatened its neighbor is thus:
21vl − vl + ε = vl + ε. (3) 20 20
And the minimum price that a lucky community would demand from its neighbor if it threatened its neighbor is:
41vl −vh +ε=−19vl +ε. (4) 20 20
Equations (3) and (4) indicate that while an unlucky community must be bribed by its neighbor to prevent war, a lucky community is willing to bribe its neigh- bor to prevent war. Since only a lucky community will pay any extortion price if threatened, a community knows that, conditional on its neighbor’s willingness to pay any extortion price, its neighbor must be lucky. Thus, to be willing to leave its neighbor in peace, an unlucky community requires an extortion price greater than its expected profit from war with a lucky one:
vh +vl −c−vl = 11vl +ε. (5) 2 20
2222 20

144 Leeson
Table 1. Unlucky community’s expected payoffs from threatening vs. doing nothing.
Threaten
Do nothing
1􏰀 11v 􏰁 21v
Threaten
Do nothing
vl + l + ε =
20 2 2 2 20 20
21v
l + ε
1􏰀v +v 􏰁
l l − c +
l + ε
21vl vl 20
Table 2. Lucky community’s expected payoffs from threatening vs. doing nothing.
Threaten
1􏰀 11v 􏰁 1 vh− l−ε + [vh]
Do nothing
1􏰀v+v 􏰁 1􏰀 11v 􏰁
Threaten
Do nothing
l h−c +  vh+
2 20 2 2 2 2 20
l+ε
= 109vl −ε 40
49vl − ε 20
= 51vl +ε 20
vh
At this extortion price, lucky communities always surrender when threatened (vh − [11vl/20] − ε > 41vl/20). Unlucky communities never do. Any community that threatens its neighbor therefore demands an extortion price at least equal to that in Equation (5).
It is now straightforward to determine when communities will threaten their neighbors versus when they will do nothing. Table 1 presents an unlucky commu- nity’s expected payoffs for threatening its neighbor versus doing nothing when its neighbor threatens it or does nothing. Table 2 presents the same information for a lucky community.
An unlucky community has a dominant strategy: threaten its neighbor. A lucky community’s payoff-maximizing strategy depends on what strategy it expects its neighbor to follow. It is straightforward to solve for the critical probability with which a lucky community must believe its neighbor will threaten it such that threat- ening its neighbor is a dominant strategy. Where p ∈ [0, 1] is the probability that a
Human Sacrifice
lucky community assigns to its neighbor doing nothing we have: 􏰀49v 􏰁 􏰀51v 􏰁 􏰀109v 􏰁
Thus if a lucky community assigns a greater than (1 − p =) 4/15 probability to its neighbor threatening it, its dominant strategy is to threaten its neighbor. Since unlucky communities always threaten their neighbors, and there is a 50% chance that a lucky community’s neighbor will be such a community, the probability that a lucky community assigns to its neighbor threatening it can never be less than 1/2. This is greater than 4/15. Lucky communities therefore always threaten their neighbors.
Both lucky and unlucky communities have dominant strategies, so neighbors’ interactions have a unique equilibrium. All communities threaten their neighbors. Unlucky communities earn:
p[vh] + [1 − p] l − ε = p l + ε + [1 − p] l
20 20 40 15
1􏰀v+v 􏰁 1􏰀 11v 􏰁 21v
l l−c+ vl+ l+ε= l+ε.
11 ⇒ p = .
145
2 2 2 20 20
And lucky communities earn: [109vl/40] − ε. On average a community’s expected period payoff is therefore:
1 􏰀21v 􏰁 1 􏰀109v 􏰁 151v
l +ε + l −ε = 220 240 80
l. (6)
2.2 Property Protection through Human Sacrifice
Perhaps counter-intuitively, if communities destroy part of their wealth, they can do better than this. To deter its neighbor’s property aggression, each period, each community needs to destroy enough of its wealth, D, to render war with it less profitable than peace for unlucky communities:
1 􏰀 vl + vl − D 􏰁 1 􏰀 vh + vl − D 􏰁 vl
2 2 + 2 2 − c < vl ⇒ D = 10 + ε.
On average a community’s expected period payoff is therefore:
vh + vl − D = 19vl − ε. (7) 2 10
Equation (7) is greater than Equation (6). By reducing its expected wealth to just below that level which attracts an unlucky neighbor’s aggression, on average

146 Leeson
a community that destroys D saves more in property enforcement costs through violent conflict than it forgoes in enforcing its property rights by destroying wealth. Wealth destruction is efficient.
Not all methods of wealth destruction are equal, however. To effectively protect property rights wealth destruction must satisfy three conditions:
1. Wealth destruction must be preemptive: it must occur before the second stage of communities’ interactions when communities choose whether to threaten their neighbors or do nothing. If a community waits to destroy D until after the second stage of communities’ interactions, its destruction is too late to prevent its neighbor’s extortion or violent attack if its neighbor turns out to be unlucky.
2. Wealth destruction must be public and verifiable: wealth destruction only depresses a community’s expected wealth in its neighbor’s eyes if its neighbor knows the community has in fact destroyed D. And its neighbor can only know this if the community’s destruction is public and verifiable.
Because destruction is preemptive, at the time they destroy wealth, communities do not yet know which one of their society’s other communities will be the neigh- bor they ultimately confront. Thus a community’s destruction must be public to all other communities. Short of every other community traveling to a destroying community to observe its destruction directly, which is costly, wealth destruction must exhibit publicity sufficient to communicate news of destruction even to those who do not observe it directly.
Wealth destruction must also be verifiable. Certain kinds of wealth destruction can be faked. For example, while pretending to burn a mound of crops worth D, a community can burn a mound of foliage covered with a veneer of crops worth only a fraction of D. Short of inspecting the mound to ensure it consists of wealth amounting to D, which is also costly, wealth destruction must be difficult to fake such that D’s destruction is verifiable.
3. Contributions to wealth destined for destruction must be incentivized: the indi- vidual members of a destroying community aren’t indifferent to whose wealth is surrendered for destruction. Such destruction benefits the community. But the persons whose wealth is destroyed for this purpose alone bear the cost of providing that benefit. To overcome this collective-action problem, contributions of wealth for destruction must be incentivized such that making them yields individual contributors’ significant private benefits.
While many methods of wealth destruction can be preemptive, satisfying condi- tion 1, and many can be privately incentivized, satisfying condition 3, few methods
Human Sacrifice 147
of wealth destruction are effective public meters of such destruction — i.e., exhibit publicity sufficient to communicate news of destruction to those who don’t observe it and can’t be faked — satisfying condition 2. Human sacrifice is one of them.
Unlike burning a heap of corn, immolating a live human is nearly impossible to fake. Further, while news that a community burned a heap of corn may reach one or two other communities, it is unlikely to reach a significant proportion of other communities in a large society. In contrast, news that a human victim has been ceremonially immolated is far more likely to reach many other communities. Human sacrifices are spectacles and thus noteworthy.
To effectively protect property rights, wealth destruction via human sacrifice must satisfy two additional conditions:
4. Human sacrifice must destroy valuable property: to depress a community’s expected wealth sufficiently, immolating persons must destroy the requisite part of the community’s property value. This precludes immolating crimi- nals or enemies, whose death would not depress an immolating community’s expected wealth and may even raise it. To make sacrificing humans an effective vehicle of wealth destruction, a community must convert property that con- tributes to its wealth into persons who it destroys. Purchasing innocent humans with valuable property, such as part of a community’s land or that land’s output, and then slaughtering them accomplishes this.
5. Sacrificed humans must be purchased from outsiders: a sacrificing community does not care who it purchases sacrificial victims from. Such purchase and then slaughter depresses the community’s expected wealth, protecting it against prop- erty aggression. However, only persons who lie outside the society the community inhabits will be willing to sell it victims for immolation. Persons who inhabit the same society must themselves depress their expected wealth to prevent predation and thus will be unwilling to accept payments from sacrifice purchasers, which would add to their wealth rather than reduce it.
The foregoing features that wealth destruction in general and human sacrifice in particular must exhibit to effectively protect property rights generate corresponding predictions for my theory of human sacrifice. If human sacrifice is a rational tech- nology of property protection, it should exhibit each of these specific features. In Section 4, I test these predictions using human sacrifice as practiced by the Konds. First I introduce the Konds and describe their institution of immolation.
Before doing so, however, it bears emphasis that the foregoing predictions are not the only ones my theory of human sacrifice generates. For example, my theory
148 Leeson
also suggests that when relatively favorable agricultural shocks are stronger — i.e., produce relatively more output — communities should destroy more wealth and thus should sacrifice more innocents. Unfortunately, evidence from the Konds that could illuminate this prediction is unavailable. Thus the evidence from the Konds I consider later test my theory imperfectly. This points to the importance of exercising caution when drawing conclusions from the historical episode I consider.
3 Human Sacrifice among the Konds
3.1 Kond Political Economy
The Konds are an indigenous Indian people located in the eastern province of Orissa.6 This article describes these people in the first half of the nineteenth cen- tury.7 The Konds inhabited the Eastern Ghats — the intermittent mountain range that cuts across eastern India, running through Orissa.8 This region overlapped the Madras and Bengal Presidencies.
The Kond economy was a primitive agricultural one. It produced rice, oils, millets, pulse, fruit, tobacco, tumeric, and mustard. Konds also raised buffaloes, goats, pigs, and poultry (Macpherson, 1865, p. 63).
Despite its primitivity, Kond agriculture was capable of “resulting in no small share of rural affluence” (C.R., 1846a, p. 49). The extent to which a Kond com- munity enjoyed such affluence depended on “the capricious climate,” most notably local weather and animal activity, which varied across communities from year to year (van den Bosch, 2007, p. 203). The fortune of a good season produced com- parative bounty for lucky communities. The misfortune of a bad season produced comparative deprivation for unlucky ones.
The Kond population consisted of several hundred thousand people.9 This pop- ulation was divided into a large number of communities. Each community was
6 On Orissa and its history, see Stirling (1904).
7 The Konds were not a homogenous people, displaying variation in language, customs, and so
on throughout the region they inhabited. Unavoidably, I abstract from these variations in my discussion. For a discussion of the people of the Orissa hills in the twentieth century, see, for instance, Bailey (1957, 1960), Banerjee (1969), and Boal (1997).
8 The Konds are sometimes divided into two groups: Sasi Konds, who inhabited the plains below the Ghats, and Maliah Konds, who inhabited the hill tracts in the Ghats. When this article refers to Konds, it refers to the latter.
9 In 1891 this population was estimated to be 620,000 (Thurston, 1909, p. 358). Though not all of these persons inhabited the hill tracts and not all of them formerly practiced human sacrifice.

Human Sacrifice 149
composed of smaller sociopolitical units, the smallest and most elemental of which was the Kond family.
The head of each Kond family was the patriarch who governed his family mem- bers’ activities. A group of families—connected by real or fictitious lineage — formed the next Kond sociopolitical unit, the village, headed by the lineage patri- arch. A collection of adjacent villages in turn formed a mutah, consisting of different branches of a Kond tribe or clan. Each mutah had a leader regarded as the head of the family that first took possession of the land the mutah occupied. Finally, a collection of contiguous mutahs formed a Kond tribe headed by a leader called a bisaye.10
The typical Kond community — i.e., group of cooperating Konds — was the mutah or tribe, though in some cases a community could be an independent village, such as when all of a tribe branch’s members inhabited the same settlement. In other cases, but less frequently, a Kond community could encompass the members of multiple tribes, such as when branches from different tribes, or multiple tribes, allied with one another.
Kond political organization was informal. The authority of the leaders described above rested “on moral appliances as contra-distinguished from coercive or forcible measures” (C.R., 1846a, p. 36). Kond country was anarchic: it “had no traditions of . . . centralized government” (Bailey, 1957, p. 181).11
Moral appliances secured governance at Kond society’s lowest levels — within communities. But they failed to do so at its higher levels — those involving inter- actions between communities. As Macpherson described it, “The institutions of the Khonds suffice, generally, to maintain order and security within tribes, but no gen- eral authority exists to control these, or their branches . . . to determine questions of civil right between them, and to enforce its decisions” (Selections from the Gov- ernment of India, 1854, p. 57). “Hence,” he concluded, “while within each tribe
10 A bisaye was typically not a Kond but rather an Oriya. Oriyas were Hindu people originating from the lowlands, some of whom historically settled among the Konds. Macpherson (C.R., 1856a) and Mullens (1854) describe the presence of an additional Kond leader — a “federal patriarch” — whose purview was to oversee relationships between several allied Kond tribes. They refer to this person as the bisaye. Others refer to tribal heads with this term (see, for instance, Campbell, 1864; Maltby, 1882).
11 Kond territory was divided into numerous zemindaries officially overseen by rajas located in the plains. However, rajas performed no governance functions for the hill Konds and, apart from paying nominal alligiance to their rajas when it suited them, the Konds lived independently of the rajas’ control (see, for instance, Bailey, 1957, p. 162; Padel, 2000, p. 15).

150 Leeson
order and security prevail, beyond all is discord and confusion” (Macpherson, 1865, p. 81).12
A chief manifestation of this discord and confusion was inter-community prop- erty encroachments. Kond communities had no compunction about attacking other communities to seize their property when that property was sufficiently attractive relative to their own, such as when agricultural output variation produced wealth disparities. “Their favorite phrase is that they ‘seize whatever they like best;’ which happens unfortunately to be . . . just those things which their neighbours also most esteem — property of the solid kind — cattle, rice, and implements of husbandry” (C.R., 1846a, p. 41).
In the absence of government, Kond communities had but one means of protecting their property against their neighbors’ aggression: defensive warfare.13 Communities built defensive structures as their technology and the hilly terrain allowed, such as forts and rude stone walls. But “otherwise,” a British officer observed, Kond communities remained “wholly unprotected” (Frye, 1860, p. 7). Even in warfare, which Konds fought with slings, bows, and axes, defensive technologies were absent. Shields, for example, were unknown to them. “Torn and distracted by interminable feuds and sanguinary quarrels, which [the Konds had] no means of adjusting, except by farther unavailing violence and bloodshed,” inter-community war was frequent (C.R., 1847, p. 28).14
3.2 Kond Sacrifice
Kond communities sacrificed humans.15 Their victims were called meriahs. Konds purchased these persons from meriah sellers called Doms (or Pans) who I discuss
12 Exacerbating the difficulty of securing cooperation between communities was the presence of multiple languages among the Konds — one community’s language sometimes being unintelli- gible to others. According to Macpherson, “A Khond of one district,” for instance, “has been found unable to hold communication with one of a neighboring tribe” (C.R., 1846a, p. 29). Communication-exacerbated discord in turn contributed to a situation in which there was “but little intercourse” between groups, which made communication difficulties that much worse, contributing further to discord, and so on in a vicious cycle (Campbell, 1861, p. 86).
13 For a description of a ritualized form of Kond warfare, see Macpherson (1865, p. 80).
14 As Macpherson (1865, p. 81) noted, “hostility is limited or modified by special compacts; but
war is still the rule, peace the exception.”
15 While human sacrifice was a near-universal practice among the Konds, it was not totally
so. Within Kond country there were several smaller areas in which human sacrifice was not practiced. In some of these areas female infanticide existed. In the others neither sacrifice nor infanticide existed. See, Selections of the Government of India (1854, pp. 54–55).

Human Sacrifice 151
in Section 4.16 In principle meriahs could be persons of any age, sex, race, or caste. In practice they were nearly always non-Konds.17
“The meriah sacrifice was an elaborate affair” (van den Bosch, 2007, p. 214). It involved rituals for preparing meriahs for immolation, rituals for the immolation itself, and rituals following the immolation. Sacrifice was “celebrated as a public oblation . . . both at social festivals held periodically, and when special occasions demand[ed] extraordinary propitiations” (Macpherson, 1865, p. 113).
Every community held at least one of these festivals every year. Typically a single meriah was sacrificed at each festival. But this was a lower bound. Kond- country visitors occasionally reported sacrifices of upwards of 20 meriahs at a time (Selections from the Government of India, 1854, p. 22). The general impression of British officers who visited Kond country was that “the number of Meriahs annually immolated” was large — very large — indeed, “far larger than could readily be credited” (Selections from the Government of India, 1854, p. 28; see also, C.R., 1846a, p. 61).18 “One thing is painfully certain, and that above the possibility of question, that the number is great beyond what any humane spirit can contemplate without the thrill of horror” (C.R., 1846a, p. 61).
Immolation festivals were large, raucous, three-day parties at which attendees engaged “in the indulgence of every form of wild riot, and generally of gross excess”
16 As I discuss below, Konds almost always purchased meriahs from these persons. However, in rare cases Konds sold their children to other Konds as meriahs. Two circumstances seem to have given rise to such an occurrence. First, the destitution of the Kond seller, who could not survive without making the sale. And second, the efforts of the British to suppress sacrifice, which made it harder for Pan middlemen to operate, in turn forcing Konds to purchase persons from the hills for sacrificial purposes. See C.R. (1848, p. 311).
17 They were also “generally in innocent childhood” (Macpherson, 1843, p. 199; see also, Selections from the Government of India, 1854, p. 17).
18 As Macpherson lamented (1843, p. 180), “the number of sacrifices in a Khond district depends on circumstances so variable, that it is scarcely possible to form an estimate . . . of their annual average” (see also, Campbell, 1864, p. 267). But this did not stop officers from offering esti- mates, which, unsurprisingly, vary greatly. Campbell (1864, p. 268) attributed an estimate of 500 to Macpherson and himself offered an estimate of 150. A British survey officer “estimated the number of victims who are to be put to death in the forty Mootahs of Ganjam alone, at 240; but these Mootahs,” he hastened to add, “are but a small part of the wide region over which we know that this practice prevails” (Macpherson, 1865, p. 151). Another British officer suggested that “A sacrifice is never offered in any village oftener than once in 12 years” (Selec- tions from the Government of India, 1854, p. 10). Perhaps the only communities he was aware of were composed of more than 12 villages. In any event, his statement is not supported by any other observers, who describe sacrifice as very frequent. As Boal (1997, p. 100) summarized it echoing Macpherson, while “the number of sacrifices in a district were dependent on circum- stances too numerous and variable to estimate the annual average, . . . they were certainly a very great number.”

152 Leeson
(Macpherson, 1865, p. 118). The villages that composed each community took turns sponsoring the festival—purchasing the meriah and hosting the party. The festival’s other attendees were representatives from other villages who came to celebrate the rite.
These festivals’ main event was the immolation itself, which took place on the party’s third day. On this day the sponsoring Kond village’s head brought the meriah, intoxicated with alcohol or opium, to a spot previously appointed for the sacrifice. There the festival’s participants took part in prayers and supplications to a malevolent earth goddess named Tari Penu to whom the meriah was being offered. The priest who led these prayers, the village head, and the meriah’s seller, who also attended the festival, then engaged the meriah in a lengthy, formulaic, ritual “conversation” carried on, on the intoxicated victim’s part, by proxy.
In some cases the victim’s arms and legs were broken to prevent his motion. After this and some final prayers, the priest gave the word, and “the crowd throws itself upon the sacrifice and strips the flesh from the bones, leaving untouched the head and intestines” (Macpherson, 1865, p. 128). While cutting the victim to pieces in this fashion was common, Konds sometimes used other modes of immolation — all of them spectacular and spectacularly brutal — ranging from drowning the victim in a pit of pig’s blood to beating him to death with brass bangles, always followed by cutting him into small pieces (see, for instance, Campbell, 1861, pp. 30–31; Campbell, 1864, p. 184; Macpherson, 1865, p. 130; Maltby, 1882, p. 87; Frazer, 1922, p. 419).
The meriah thus slaughtered, the festival reached its crescendo. The chief gave a pig or buffalo to the priest and the meriah’s seller, concluding the event. Each of the participating villages’ representatives took a strip of the corpse’s flesh and departed for their settlements where they shared it with their village members who buried the flesh in their fields.19
4 Testing the Theory of Human Sacrifice
The theory of human sacrifice developed in Section 2 predicts that human sacrifice should exhibit specific features. Evidence from the Konds allow me to test several of these predictions. The evidence supports them.
19 Several days later a buffalo was slaughtered and feasted on by the festival’s attendees. And a year following the immolation festival a pig was slaughtered.

Human Sacrifice 153
4.1 Human Sacrifice Destroys Valuable Property
According to Kond belief, “To be acceptable to Tari a victim had to be purchased” by a sacrificing community (Maltby, 1882, p. 79).20 As Macpherson put it, “in all cases, it must be bought with a price—an unbought life being an abomination to the deity” (C.R., 1846a, p. 61). “This condition,” Campbell (1861, pp. 26–27) emphasized, “is essential” (see also, Elwin, 1944, p. 54).
Kond communities made these purchases by trading wealth — land, agricultural output, work animals and tools used to produce that output, and goods, such as metal ornaments and vessels, obtained with that output — for meriahs. In slaugh- tering meriahs, communities destroyed the valuable property they exchanged for them.
Konds’ unit of account was an article of such property they called a “life” (or gonti). A life consisted of property such as “a bullock, a buffalo, goat, a pig or fowl, a bag of grain, or a set of brass pots . . . . A hundred lives, on average . . . consist[ing] of ten bullocks, ten buffaloes, ten sacks of corn, ten sets of brass pots, twenty sheep, ten pigs, and thirty fowls” (Macpherson, 1865, p. 64). Meriah prices were rendered in these units. And their prices were considerable (Padel, 2000, p. 134). According to Campbell (1861, p. 28), a single meriah cost a purchasing community “from ten to sixty” lives. This constituted a “very great expense attendant upon procuring the victims” for sacrifice (Campbell, 1864, p. 268).
The importance Konds attached to the necessity of expending large sums to procure meriahs, and as I discuss below, publicizing those expenditures, is evidenced in the ritual dialogue between a sacrificing community’s priest, or jani, the presiding chief, and the meriah (via proxy) that Konds acted out at immolation festivals. This lengthy dialogue publicly declared the fact that the community had purchased the meriah about to be slaughtered with its valuable property. Consider the following extracts from such a dialogue as recounted by Macpherson (1865, pp. 123–128; see also Campbell, 1861, pp. 30, 139; Selections from the Records of the Government of India, 1854, p. 4; Thurston, 1912, p. 204):
[After initial ritualized dialogue from the chief and priest] the Victim answers — ‘Have you no enemies, no vile and useless child, no debtor to another tribe, who compels you for his debts to sell your lands; no
20 An acceptable meriah could also be one that was “born a victim, that is, the son of a [purchased] victim father; or if he was devoted as a child to the gods by his father or natural guardian” (Macpherson, 1865, p. 115). Born victims were often allowed to live and work within the Kond community that owned them, sometimes reaching adulthood before eventually being sacrificed.

154
Leeson
coward, who in time of battle skulks with another tribe? Have you none of these to seek out and sacrifice?’
The Janni replies — ‘We have acted upon quite different views. We did not kidnap you on the road, nor while gathering sticks in the jungle, nor when at play.... Such sacrifices would be of no avail. To obtain you, we cleared the hills and the jungle, fearless of the tiger and the snake. We stinted ourselves to fill your parents, and gave them our brass vessels....’
The Victim. — ‘And did I share the price which my parents received? Did I agree to the sale?.... You, O my father, and you, — and you, — and you, — O my fathers! do not destroy me!’....
[The chief declares]: ‘This usage is delivered down to us from the first people of the first time.... O child! we must destroy you.’....
[After further dialogue, the chief continues]: ‘We cannot satisfy you. Ask your father [i.e., seller], who is present. I satisfied him with my favourite cattle, my valuable brass vessels, and my sheep, and with silken and woollen cloths, and axes. A bow and arrows, not four days old, I have given to his fancy’....
[More dialogue, then the priest repeats]: ‘by our cattle, our flocks, our pigs, and our grain we procured a victim and offered a sacrifice.’
The purchase theme was repeated a final time “When the axes of the crowd [were] raised to complete the rite,” the community’s members shouting as they lurched at the victim to shred him: ‘We bought you with a price!”’ (Macpherson, 1843, p. 199).
Since “It was necessary that the victim should always be purchased,” as the ritual dialogue above intimates, “the Khonds did not sacrifice criminals, or prisoners captured in war” (Maltby, 1882, p. 83; see also, Thurston, 1912, p. 200). Such persons were “not considered fitting subjects” to satisfy Tari according to Kond belief (Selections from the Records of the Government of India, 1854, p. 4).
4.2 Wealth Destruction is Preemptive
Each Kond community sacrificed humans during a fixed part of the year correspond- ing to the agricultural cycle. As Macpherson (1852, p. 213) observed, “The periodic common sacrifices are generally so arranged by tribes and divisions of tribes, that each head of a family is enabled, at least once a year, to procure the shred of flesh for his fields, and usually about the time when his chief crop is laid down.” Some
Human Sacrifice 155
communities also sacrificed at “about the time of the . . . harvest” (Macpherson, 1865, p. 107). Finally, “it [was] considered in the last degree desirable that several offerings, according to the promise of the year, should intervene betwixt” these end points of the growing cycle (C.R., 1846a, p. 60; see also, Mullens, 1854, p. 152).
By sacrificing humans between the sewing and harvesting of crops, Kond com- munities destroyed wealth preemptively. Communities did not learn their nature- assigned output values, and so could not sensibly choose whether or not to aggress against their neighbors’ property, until the harvest, and thus agricultural season, was completed. By sacrificing humans during the agricultural cycle but not appre- ciably after its completion, Kond communities destroyed wealth before other com- munities realized their output values, and in the event that those values incentivized aggression, before communities could mobilize for such aggression.
4.3 Wealth Destruction is Public and Verifiable
Kond immolation festivals were public events, news of which traveled far and wide, reaching those who did not attend them. According to Macpherson (1843, p. 182), “From these festivals no one is excluded.” And as Campbell (1864, p. 53) described, “The sacrifice, to be efficacious, must be celebrated in public before the assembled people.”
The public assemblies before which meriahs were spectacularly sacrificed were populous, being “generally attended by a large concourse of people” who came to participate in the dramatic rite (Macpherson, 1843, p. 182; see also, Davies, 1981, p. 79). Thus when a British officer arrived at one Kond community in the course of its immolation festival, he observed “The Khonds from the neighbouring villages [who] . . . kept arriving in multitudes” (Selections from the Government of India, 1854, p. 26).
Open, well-attended, public immolation festivals ensured that a sizeable audi- ence observed an immolating community’s sacrifice. Still, given the cost involved, not every Kond community or every one of such a community’s members could attend every other community’s immolation festival. For those that were not able to attend, Konds’ sacrificial ritual used an additional mechanism to ensure the sacri- fice’s publicity. As Campbell described it, after the meriah was slaughtered (Risley, 1892, p. 406),
The flesh cut from the victim was instantly taken home by the persons who had been deputed by each village to bring it.... In each village all who stayed at home fasted rigidly until the flesh arrived. The bearer
156
Leeson
deposited it in the place of public assembly, where it was received by the priest and the heads of families. The priest divided it into two portions, one of which he offered to the earth goddess by burying it in a hole in the ground with his back turned, and without looking. Then each man added a little earth to bury it, and the priest poured water on the spot from a hill gourd. The other portion of flesh he divided into as many shares as there were heads of houses present. Each head of a house rolled his shred of flesh in leaves, and buried it in his favourite field.
In this way a sacrificing community’s public ritual was extended via a “piggy- backed” ritual performed by an immolation festival’s external attendees — itself public in each of their villages and carried out “with great pomp” — extending knowledge of the sacrificing community’s immolation to the inhabitants of villages who were not themselves able to participate (Eliade, 1958, p. 345).
Kond communities’ wealth destruction via human sacrifice also overcame the problem posed by the possibility of faking wealth destruction. The reason for this is simple: it is nearly impossible to fake the immolation of a live human — particularly through forms of immolation as brutal as those the Konds used.
Nevertheless, a difficulty remained. How could immolation-festival attendees, and thus those who heard about such festivals second-hand, be sure that the victim a community immolated was a purchased victim? Further, how could other commu- nities have an idea of how much wealth an immolating community destroyed when it sacrificed such a victim?
Kond communities ensured others that the meriahs they sacrificed were pur- chased by leveraging the persons who sold them meriahs. Recall that a meriah’s seller attended the immolation festival of the community he sold to. His presence acted as a kind of sales receipt, verifying the fact that the meriah about to be sac- rificed had indeed been bought. Of course, it was possible for a Kond community to attempt to bribe such an individual to verify a sale that never happened. But the bribe a seller would have required to do this exceeded the wealth a community needed to destroy to protect its property. Because of this, communities preferred to buy their meriahs instead.
To protect its property rights through wealth destruction, a community needed to destroy D. This reflects the price a community would spend on a meriah. However, to lie on a community’s behalf, a meriah seller would demand the difference between the community’s expected payoff under property protection and its expected payoff under property aggression from Equation (6), since this is what the community would lose if the seller “outed” its attempt to bribe him by supplying evidence that
Human Sacrifice 157
it tried to do so. This difference is larger than D: 2vl − [151vl/80] > [vl/10] + ε. Because a community earned more by purchasing its meriah than by bribing a meriah seller to lie about having made it a sale, a seller’s presence at a commu- nity’s immolation festival reliably verified the fact that the community had indeed purchased the victim it was about to slaughter.
Knowing a meriah had been purchased was not the same as knowing how much that meriah had been purchased for, however, and thus how much wealth an immo- lating community destroyed by sacrificing him. Konds addressed this issue in two ways. First, as themselves buyers in the meriah market, communities knew the approximate price of a meriah. Second, in some areas at least, they knew the pre- cise price of a meriah, as there was “a fixed price for each person” set by custom (Selections from the Government of India, 1854, p. 10).
4.4 Contributions to Wealth Destined for Destruction are Incentivized
The typical Kond community had two sorts of “individual members”: the villages that composed the community and the families that composed those villages. It was therefore necessary to motivate each village to contribute wealth for destruction and, within each village, to motivate individual families to do so.
One means of achieving this, and one on which Kond communities in part relied, was simple turn taking. As noted above, each of the villages composing a Kond community took turns sponsoring the community’s immolation festival. According to Macpherson (1865, p. 113), “When a tribe is composed of several branches, the victims for the fixed offerings are provided by the branches in turn, the cost being defrayed by contributions borne by each person according to his means.”
While helpful, this solution to Kond communities’ internal collective-action problem was imperfect. The problem that remained is the one alluded to by the second part of Macpherson’s statement. Within each village in its turn for sponsoring sacrifice, each family contributed to the village’s purchase “according to his means.” Differences in families’ means resulted from differences in their outputs.
Although minimal compared to the output variation that existed between com- paratively geographically separated communities, output varied between families in the same village and between the villages that composed the same community. Natural differences in geography, the effects of even the same weather, and differ- ences in individuals’ productivity within the same village and between villages in the same community ensured as much.
158 Leeson
Intra-community output variation reintroduced the possibility for free riding on contributions of wealth destined for destruction. By misrepresenting what nature had provided it, a village, or family in such a village, could plausibly claim that it was not in a position to contribute the wealth required to purchase a meriah in a particular cycle, shirking on its obligation to make a contribution. Konds’ solution to this problem was the second means they employed to overcome the collective- action problem that threatened their ability to use wealth destruction to protect property rights: human sacrifice was rendered a religious obligation.
Konds believed their fate rested in the hands of Tari Penu — the malevolent earth goddess to whom they offered meriahs.21 To “obtain abundant crops, to avert calamity, and to insure prosperity in every way” they required her favor (Camp- bell, 1861, p. 25). Tari craved the blood of sacrificial human victims and “caused all kinds of afflictions and death if she was not satisfied,” most notably “through war and natural calamities” (van den Bosch, 2007, p. 224; Das, 1986, p. 5). By religionizing immolation this way, Konds rendered “Contribution to the support of the ceremonial of human sacrifice, the rite of which constitutes the public reli- gion . . . an indispensable condition of association in a Khond tribe” (Macpherson, 1843, p. 178).
According to Kond belief, the positive results of satisfying the earth goddess and negative results of denying her could be felt not only collectively, but also individually, at the village level or that of villages’ individual families. Indeed, according to Kond religion, to avert private calamities, such as illness or the death of one’s animals, individual community members had to satisfy Tari’s blood thirst with human sacrifices just as villages and communities had to do so to prevent social calamities. As Macpherson described it (C.R., 1846a, pp. 60–61):
The private performance of bloody sacrifice is deemed necessary, when any extraordinary calamity marks the anger of the deity towards a par- ticular house, as, for example, when a child, watching a flock, perishes by a tiger — the form of which is believed to be assumed by the Earth- goddess for the purposes of the wrath. On application to the priest, he of course refers the visitation to the neglected worship of the dread deity, and generally demands an immediate victim. If this requisition
21 For an extensive discussion of Kond religious beliefs, which displayed some heterogeneity, see Macpherson (1852).

Human Sacrifice 159
cannot be complied with, a goat is led to the place of sacrifice, where its ear is cut off and cast bleeding upon the earth — a pledge that must be redeemed by human blood, at whatever cost, within the year.
Free riding on others’ contributions to meriah purchases exposed the free rider to Tari’s wrath and, in consequence, misfortune. Fear of such wrath privately incen- tivized each of a Kond community’s individual members to contribute “according to his means” to his community’s meriah purchases.
4.5 Sacrificed Humans are Purchased From Outsiders
Kond communities purchased meriahs from the members of an untouchable caste called Doms (or Pans). Many Doms lived outside Kond country in the foothills of and the plains below the Ghats. Other Doms lived inside Kond country, but in separate, itinerant communities on the outskirts of Kond communities. “Being outcast, they are never allowed to live within a village, but have their own little hamlet adjoining a village proper” and “frequently change their place or residence and their protectors” (Fawcett, 1901, p. 34; C.R., 1846a, p. 47).
The latter Doms had close connections to the former — their families and extended social groups inhabiting the foothills and plains. “They keep up constant intercourse and connection by marriage, with the families of their race who live in the low country near the ghauts” (Selections from the Government of India, 1854, p. 50).
Kond communities used Doms for two purposes: (1) to trade with persons located outside Kond country and (2) to perform ritually impure acts forbidden to Konds according to their religion, but permissible to already ritually polluted outcastes such as the Doms. In the former capacity, Doms operated as middlemen connecting Konds in the hill tracts and non-Konds in the plains, exchanging Kond produce, such as tumeric, for that of the plains inhabitants, such as salt, salt fish, and cloth. Indeed, Doms “manage[d] the whole commerce of the hills” (Selections from the Government of India, 1854, p. 50; see also, Bailey, 1960, pp. 134–135). In the latter capacity, Doms handled, for instance, various Kond death rites that were taboo for Konds.
Among the middleman functions for which the Konds employed Doms was the procurement of victims for human sacrifice. As Macpherson observed, meriahs were “provided by a class of Hindu procurers, called ‘Panwas.”’ The meriah trade was simple. Doms “purchase[d] them without difficulty on false pretenses, or kidnap[ped]
160 Leeson
them from the poorer classes of Hindus in the low country” (C.R., 1846a, p. 61).22 Having done so, they traded their human victims to Kond communities in the hill tracts for wealth in the forms described earlier.
While both Doms living inside and outside the Kond society acted as meriah traders, in general it seems to have been the latter, “the inhabitants of the low countries who provided the means of carrying on the horrible practice” (C.R., 1846b, p. 54). As Campbell observed, “the parties engaged in the barbarous traffic of providing victims” were “from all the information I can gather . . . for the most part inhabitants of the low country” (Selections from the Government of India, 1854, p. 12). Because these Doms lived outside Kond society, when they received Kond wealth in exchange for meriahs, that wealth exited Kond society with them.
Even when Konds relied on Doms who lived inside their communities to pur- chase meriahs, much of the Kond wealth received as payment must have similarly exited Kond country. Dom traders in Kond country relied on associates outside it in the plains to obtain meriahs that the former could ultimately sell to Konds. As Macpherson put it, “it is certain, not only that the other Panwas, besides those who are permanently associated and identified with the sacrificing Khond tribes, provide victims, but that these are most generally procured in the first instance, by Panwas of the low country” (C.R., 1847, p. 7). Unless the latter Doms were per- forming this service for free, a significant part of meriah receipts must have flowed to these outsiders too.
4.6 Human Sacrifice Ends when Government Protects Property Rights
The theory developed in Section 2 assumes that government does not exist to protect communities’ property rights. If it did, there would be no reason to destroy wealth, and thus no reason for human sacrifice, since communities could enjoy property protection vis-a`-vis one another without this costly practice. My theory therefore delivers one final testable prediction — one undiscussed in Section 2: Kond communities should have ceased to sacrifice humans when government became available to them to protect their property rights.
The initial attempts of the British to “civilize” the Konds took two forms: threatening the Konds with violent punishments if they refused to abandon human
22 Doms also sometimes sold their own children for this purpose and, when they did not, some- times represented the persons they sold as such.

Human Sacrifice 161
sacrifice, and attempting to reason with the Konds by “educating” them about the barbarity of the practice and the scientific baselessness of their belief in the neces- sity of satisfying an earth goddess with the blood of purchased persons to ensure prosperity and prevent calamity.
Neither strategy proved effective. The British made periodic forays into Kond country during the immolation season and, when Kond communities refused to deliver up the meriahs they were bent on sacrificing, used violence against the Konds. But the communities kept immolating. Attempting to reason with the Konds proved scarcely more helpful. To appease their British visitors, some Konds sur- rendered a portion of the meriahs they intended to sacrifice. But once the British returned to the plains, they immolated the ones they held back.
Frustrated with their failure to stop Kond sacrifice, the British eventually tried another tack. Captain Macpherson, one of the officers in charge of the British agency for the suppression of human sacrifice in the hill tracts of Orissa, thought that offering a trade to the Konds might prove more effective. What, he asked, was the greatest unmet need of the Konds? What could the British offer them that they valued so highly they might be willing to give up human sacrifice in return?
Inter-community property protection. As Macpherson put it: “Now it appears distinctly that the great social defect for these clusters of tribes . . . is the want of a supreme controlling authority,–of a power able to arbitrate betwixt different tribes, and betwixt tribes and the zemindaries; and this want, I think, we may, by direct and indirect means, to a certain extent, supply” (C.R., 1847, p. 18). “The Khonds,” it occurred to him, “most anxiously desire of us justice — not betwixt man and man, which their own institutions can afford, but betwixt tribes and their divisions, which the authority of those institutions is too feeble to reach” (Macpherson, 1865, p. 178).
So Macpherson offered it to them — “by arbitrating, not merely between indi- viduals of the same tribe, but also between their several tribes and authorities” — provided that they agreed to abandon human sacrifice (C.R., 1847, p. 18). The Konds eagerly accepted his offer.
One mutah, for example, “promise[d] to relinquish from henceforth the rite of human sacrifice” on the condition “That they shall be received into the immediate protection of the Government, and shall always obtain justice from it” (Selections from the Government of India, 1854, p. 55). Another similarly agreed “to relin- quish the rite of sacrifice ‘upon the condition of their receiving protection and peace and justice from the Government”’ (C.R., 1847, p. 36). Soon other communities
162 Leeson
“spontaneously proffered to relinquish the sacrifice, mainly on the condition of obtaining protection and justice, and actually pledged themselves accordingly” (C.R., 1848, p. 275).23
Much to the officers’ delight, Kond communities kept their side of the bargain. When the British entered the hill tracts on regular visits, they found that most Kond communities had indeed left off the practice of human sacrifice.24 By the middle of the nineteenth century the Kond institution of human sacrifice had ended (van den Bosch, 2007, p. 197).
5 Summary and Conclusions
In my theory of human sacrifice the purchase and ritual immolation of innocent persons is a substitute technology for protecting property rights in agricultural soci- eties where conventional methods of property protection, such as government, are absent and nature-induced shocks precipitate conflict between communities seek- ing others’ wealth. Human sacrifice cannot only be rational in this context, but also efficient. When conflict is sufficiently costly, the wealth lost in violent clashes without human sacrifice exceeds that which is destroyed via human sacrifice.
Although the evidence required to test all of the predictions my theory generates is unavailable — most notably, that which could shed light on the prediction that communities should sacrifice more frequently when relatively favorable agricultural shocks are stronger — what evidence is available from the most significant ritual immolators of the modern era, the Konds of Orissa, seems to be consistent with my theory.
It is unknown when the Konds began purchasing and ritually immolating innocent persons. The Konds themselves remained unknown until the British discovered them in the early nineteenth century. According to what the Konds told the British, however, their institution of human sacrifice was age old. They had practiced it “from time immemorial” (Padel, 2000, p. 66). My theory, which
23 This response constituted an about face from the way in which the Konds initially responded to the British officers’ presence. That initial negative response is not surprising, however, given that at first the British were simply persons seeking to compel the Konds to give up human sacrifice without offering any substitute institution of property protection in return.
24 Some Kond communities slaughtered a buffalo or wild animal instead. As Russell (1916, p. 473) puts it, “Once in four of five years a buffalo is offered to the earth god, in lieu of the human sacrifice which was formerly in vogue.”

Human Sacrifice 163
points to the usefulness of human sacrifice in the context in which the Konds lived, explains the duration of this institution in their society. Kond sacrifice may have been, as the British Court of Directors called it in a letter to the Madras government, “odious and revolting” (Selections from the Government of India, 1854, p. 38). But it was also socially productive, which is why it persisted. The abrupt abandonment of Kond sacrifice when the British provided Kond communities property protection externally, and thus, according to my theory, human sacrifice was no longer useful, supports this claim.
The logic of conspicuous destruction my theory of human sacrifice describes is applicable beyond the Konds’ institution of ritual immolation and may be applied usefully to better understand a wide range of otherwise puzzling practices that appear to contradict the canonical economic assumption that more is preferred to less. The poverty displayed by some well-known groups — from Gypsies to ascetics — may reflect their members’ rational decisions to have more secure prop- erty rights in less wealth instead of less secure property rights in more wealth.
It is unnecessary, and I would argue, unhelpful, to approach the institutionalized purchase and ritual slaughter of innocent persons by abandoning rational choice theory. Not only does such abandonment leave one of history’s most well-known and intriguing institutions unexplained. It suggests that puzzling human behaviors and practices are beyond the power of economics to illuminate.
References
Ainslie, G. 2005. “You Can’t Give Permission to be a Bastard: Empathy and Self-Signaling as Uncontrollable Independent Variables in Bargaining Games.” Behavioral and Brain Sciences 28: 815–816.
Allen, D. W. 2002. “The Rhino’s Horn: Incomplete Property Rights and the Optimal Value of an Asset.” Journal of Legal Studies 31: S339–S358.
Anderson, T. L. and P. J. Hill. 2004. The Not So Wild, Wild West: Property Rights on the Frontier. Stanford: Stanford University Press.
Bailey, F. G. 1957. Caste and the Economic Frontier. Manchester: Manchester University Press.
Bailey, F. G. 1960. Tribe, Caste, and Nation. Manchester: Manchester University Press. Banerjee, S. 1969. Ethnographic Study of the Kuvi-Kandha. Calcutta: Government of India. Benson, B. L. 1989. “The Spontaneous Evolution of Commercial Law.” Southern Economic
Journal 55: 644–661.
Boal, B. M. 1997. Human Sacrifice and Religious Change: The Kondhs. New Delhi: Inter-
India Publications.
Campbell, J. 1861. Narrative of Major John Campbell of His Operations in the Hill Tracts
of Orissa for the Suppression of Human Sacrifices and Female Infanticide. London: Hust and Blackett.
164 Leeson
Campbell, J. 1864. A Personal Narrative of Thirteen Years’ Service Amongst the Wild Tribes of Khondistan for the Suppression of Human Sacrifice. London: Hurst and Blackett.
C.R. 1846a. “Goomsur; The Late War there — The Khonds or Hill Tribes.” Calcutta Review 5: 1–85.
C.R. 1846b. “The First Series of Government Measures for the Abolition of Human Sac- rifices among the Khonds.” Calcutta Review 6: 45–108.
C.R. 1847. “Captain Macpherson and the Khonds.” Calcutta Review 8: 1–51.
C.R. 1848. “The Khonds — Abolition of Human Sacrifice and Female Infanticide.”
Calcutta Review 10: 273–341.
Dalton, E. T. 1872. Descriptive Ethnology of Bengal. Calcutta: Office of the Superintendent
of Government Printing.
Das, B. S. 1986. “Prefatory Introduction to the New Edition.” In Human Sacrifices in
India, John Campbell (ed.). Delhi: Mittal, pp. 1–24.
Davies, N. 1981. Human Sacrifice: In History and Today. New York: William Morrow and
Co.
Demsetz, H. 1967. “Toward a Theory of Property Rights.” American Economic Review
57: 347–359.
Eliade, M. 1958. Patterns in Comparative Religion. New York: Sheed and Ward.
Elwin, V. 1944. “Notes on a Khond Tour.” Man in India. 24: 40–58.
Fawcett, F. 1901. “Notes on the Dombs of Jeypur, Vizagapatam District, Madras Presi-
dency.” Man 1: 34–38.
Frazer, J. G. [1922] 2010. The Golden Bough: A Study in Magic and Religion. Old
Saybrook, CT: Konecky and Konecky.
Friedman, D. 1979. “Private Creation and Enforcement of Law: A Historical Case.”
Journal of Legal Studies 8: 399–415.
Frye, J. P. 1860. “On the Uriya and Kondh Population of Orissa.” Journal of the Royal
Asiatic Society of Great Britain and Ireland 17: 1–38.
Hunter, J. M., H. M. Teaf, Jr., and A. O. Hirschman. 1957. “Discussion.” American Eco-
nomic Review 47: 57–66.
Johnsen, D. B. 1986. “The Formation and Protection of Property Rights among the South-
ern Kwakiutl Indians.” Journal of Law and Economics 15: 41–67.
Leeson, P. T. 2007. “An-arrgh-chy: The Law and Economics of Pirate Organization.”
Journal of Political Economy 115: 1049–1094.
Leeson, P. T. 2013. “Gypsy Law.” Public Choice 155: 273–292.
Macpherson, S. 1843. “An Account of the Religious Opinions and Observances of the
Khonds of Goomsur and Boad.” Journal of the Royal Asiatic Society of Great Britain
and Ireland 1: 172–199.
Macpherson, S. 1852. “Religion of the Khonds in Orissa.” Journal of the Royal Asiatic
Society of Great Britain and Ireland 13: 216–274.
Macpherson, S. 1865. Memorials of Service in India. London: John Murray.
Maltby, T. J. 1882. The Ganjam District Manual. Madras: Lawrence Asylum Press. Mullens, J. 1854. Missions in South India. London: W.H. Dalton.
Padel, F. 2000. The Sacrifice of Human Being: British Rule and the Konds of Orissa.
Oxford: Oxford University Press.
Risley, R. H. 1892. The Tribes and Castes of Bengal: Ethnographic Glossary. vol. 1,
Calcutta: The Bengal Secretariat Press.
Russell, R. V. 1916. The Tribes and Castes of the Central Provinces of India. vol. 3,
London: Macmillan.
Scott, J. C. 2009. The Art of Not Being Governed: An Anarchist History of Upland South-
east Asia. New Haven: Yale University Press.
Human Sacrifice 165
Selections from the Records of the Government of India (Home Department). 1854. History of the Rise and Progress of the Operations for the Suppression of Human Sacrifice and Female Infanticide in the Hill Tracts of Orissa. Calcutta: F. Carbery, Bengal Military Orphan Press.
Skarbek, D. B. 2011. “Governance and Prison Gangs.” American Political Science Review 105: 702–716.
Stigler, G. J. 1992. “Law or Economics?” Journal of Law and Economics 35: 455–468. Stirling, A. 1904. An Account (Geographical, Statistical, and Historical) of Orissa Proper
or Cuttack with Appendices. Calcutta: Bengal Secretariat Press.
Thurston, E. 1909. Castes and Tribes of Southern India. vol. 3, Madras: Government
Press.
Thurston, E. 1912. Omens and Superstitions in Southern India. New York: McBride, Nast
and Co.
van den Bosch, L. 2007. “Human Sacrifice among the Konds.” In The Strange World of
Human Sacrifice, Jan N. Bremmed, (Ed.). Leuven: Peeters, pp. 195–227.
}}