7 The economics of torture

1 The Economics of Torture Pavel Yakovlev [email protected] 1. Introduction Torture is an act by which severe pain (mental or physical) is intentionally inflicted on a person.1 Although torture is officially acknowledged by virtually all countries to be an extreme violation of human rights, two thirds of all countries, according to Amnesty International, do not consistently abide by anti-torture treaties.2 Despite being condemned by law and several international treaties, torture is still widely used today by many governments, many of which claim to be democratic (Neumayer 2005; Henry 1978; Rejali 2007; McCoy 2005). Existing research on torture and other human rights abuses is dominated by conflict scholars, political scientists, human rights scholars, and sociologists. While their scholarly contributions are very valuable, an economic analysis of torture would be highly complementary, but it is largely absent from the literature. This chapter offers a multidisciplinary literature review on this topic. The review begins with a historical perspective on torture followed by a discussion on anti-torture treaties, some empirical evidence on the determinants of torture, and positive arguments for or against torture. In addition, the chapter develops a simple model that examines the optimal level of torture that will be chosen when the costs of torture are widespread 1 The United Nations Convention Against Torture defines torture as any act by which severe pain or suffering, whether physical or mental, is intentionally inflicted on a person for such purposes as obtaining from him or a third person information or a confession, punishing him for an act he or a third person has committed or is suspected of having committed, or intimidating or coercing him or a third person, or for any reason based on discrimination of any kind, when such pain or suffering is inflicted by or at the instigation of or with the consent or acquiescence of a public official or other person acting in an official capacity. 2 Anti-torture treaties include but are not limited to Universal Declaration of Human Rights, U.N. Convention Against Torture, 3rd and 4th Geneva Conventions. 1 2 and the benefits are concentrated. The model shows that the optimal level of torture depends on the probability of retaliation from the opposing side in a conflict. 2. A Historical Primer to Torture For much of human history, torture has been a legitimate method of coercion, intimidation, or information gathering. Torture has been used in all societies in the ancient world with the possible exception of the Hebrew people (Evans and Morgan 1998). Often, prisoners of war were either killed or taken into slavery, which usually meant facing inhumane treatment and torture in ancient Egypt, Greece, and Rome, for example. During the Middle Ages, the incidence and severity of torture expanded beyond what the Roman law would have sanctioned (Evans and Morgan 1998). Only when the movement for the abolition of torture and for the establishment of penal reform swept throughout Europe in the eighteenth century did torture begin to be widely regarded as an unacceptable and barbaric practice (Evans and Morgan 1998). However, the moral triumph over torture did not lead to a complete elimination of torture as it re-emerged in the nineteenth and twentieth century in Austria, Italy, and Russia and later in the twentieth century in Algeria (1954-62), Northern Ireland (1963-94), Israel, and the Soviet Union (Evans and Morgan 1998). The Eastern Front of World War II was the largest single front in the history of warfare and was unparalleled in the mistreatment of war prisoners. Soviet prisoners of war, or POWs, died in large numbers in German prison camps from brutal treatment, starvation, and overall neglect. In contrast, the treatment of U.S. and other Western POWs in Nazi Germany was relatively decent and largely in accordance with the Geneva Conventions.3 Allegedly, some 3 For brevity, this chapter refers to the various articles and versions of the Hague and Geneva conventions simply as the Hague and Geneva conventions. 2 3 German prison camps even offered the Western POWs of lower rank financial compensation for their labor, while officers of higher rank were excused from work altogether (Reid, 1953). However, Nazi Germany refused to extend the same treatment to the Soviet POWs supposedly because the Soviet Union did not sign the 1929 Geneva Convention (Rolf, 1998; North, 2006). The Japanese empire did not treat prisoners of war in accordance with the Geneva Conventions either. In fact, Emperor Hirohito issued a directive that removed the constraints of the Hague Convention in order to worsen the treatment of Chinese POWs (Fujiwara, 1995). More recently, The Economist (May 17, 2008 p. 102) reported that the U.S. guards at the Abu Ghraib prison in Iraq were allegedly told by their superiors that the protections afforded by the Geneva Conventions did not apply to the detainees they were holding (Gourevitch and Morris, 2008). Although torture has been used for a variety of purposes since its inception, its application has gone through changes in recent times. In contrast to modern judicial practices, torture and punishment were often one and the same. However, the emergence of the modern state was characterized by the creation of doctrines of human rights and social contract, which made torture unacceptable (Evans and Morgan 1998). In the past, the use of torture was more likely to be linked to the status of the victim such as in the case of mistreatment of AfricanAmerican slaves in the Southern States or those considered to be “racially impure” in Hitler’s Germany and Mussolini’s Italy, and under apartheid in South Africa. Based on this evidence, Evans and Morgan (1998) argue that the use of torture seems to be inseparably linked to the changing nature of the state and its relationship with its citizens. In totalitarian states, for instance, torture is likely to be used more for intimidation purposes rather than information gathering. 3 4 3. To Torture, or not to Torture: What’s Treaty Got to Do with It? The aforementioned differences in treatment of Soviet and American POWs by Nazi Germany in WWII might suggest that the incidence and severity of torture might depend on whether a country has signed the appropriate convention. This is not necessarily so. For instance, Germany’s official justification for the poor treatment of the Soviet POWs was that the Soviet Union had not signed the Geneva Conventions. However, under article 82 of the Geneva Convention (1929), signatory countries had to give POWs of all countries, signatory or not, the same rights. Furthermore, a month after the German invasion of the Soviet Union in 1941, the USSR offered Nazi Germany the opportunity to sign a reciprocal adherence to the Hague Convention. The offer was left unanswered (Beevor, 2001). There is a growing body of evidence showing that signing various human rights conventions does not guarantee actual compliance with these treaties. On the contrary, dictatorships with higher levels of torture might be more inclined to join the U.N. Convention Against Torture or CAT (Vreeland 2008; Hathaway 2003, 2007). Some of the most repressive governments have ratified various human rights treaties, but appear to have a questionable desire to comply with these treaties in reality (Hafner-Burton and Tsutsui, 2007). Surprisingly, treaty ratification might be followed by even higher levels of torture and inhumane treatment (Hathaway 2002, 2003). There is evidence suggesting that compliance with human rights treaties and avoidance of torture are more effective in established democracies with strong civil institutions and activism (Hafner-Burton and Tsutsui, 2007). So, why do repressive governments voluntarily sign human rights treaties, such as CAT, that prohibit torture and other inhumane acts? Perhaps, signing human rights treaties allows the totalitarian regimes to create an appearance of civil rights in order to reduce the probability of conflict with powerful democratic 4 5 countries and continue to receive more international aid. In the absence of credible enforcement mechanisms to ensure that these nations actually comply with the human rights treaties, signing the treaty is a low cost option for dictators to maintain the status quo. Alternatively, by signing CAT, the oppressive governments make a small concession to internal political opposition, while still able to rely on torture and other oppressive strategies to maintain their grip on power. Hence the paradox: dictatorships sign international human rights agreements that go against their ability to maintain internal control. Neumayer (2005) conducts a quantitative analysis of whether international human rights treaties actually increase respect for human rights and finds that human rights treaties might be able to improve respect for human rights, conditional on the extent of democracy and the strength of civil society. However, somewhat disheartening evidence comes from McCoy (2005), who investigates whether or not the exposure to School of the Americas (SOA) professional military training instills a greater respect for human rights in SOA graduates. McCoy finds that while the overall number of abusers is small among SOA graduates, the abusers themselves are disproportionately represented by officers and students who took multiple SOA courses and as opposed to their counterparts who took only one course. McCoy’s findings imply that repeated exposure to SOA training is associated with increased human rights violations. One of the possible explanations for this puzzling result is that officers, who often become convenient ‘scapegoat’ targets for their superiors, tend to take more SOA courses than regular soldiers. In light of this rather ironic evidence, one may wonder about the effectiveness of anti-torture treaties in general and SOA program in particular. 4. Positive Arguments on Torture 5 6 In many present day countries, torture is explicitly outlawed by the constitution and is subject to criminal punishment. Moreover, courts almost everywhere are required to exclude evidence obtained through torture. Nonetheless, official intolerance of torture seems to diverge from practice even in countries that are often perceived as the champions of democratic values and human rights. While there are many moral arguments for and against torture, this chapter avoids normative discussions on torture in favor of the objective (positive) analysis of torture. Although generally objecting to torture, a few scholars argue that torture should be allowed under extreme or dire circumstances. They typical cite the “ticking time bomb” argument in defense of torture which refers to situations where the timely access to information obtained through torture is justified by the many lives that could be saved. In other words, their implicit cost-benefit analysis justifies torture when the value of lives saved vastly outweighs the moral cost of torture. However, Alex Tabarrok argues that the ticking time bomb argument can allow the government to abuse its torture powers.4 Tabarrok proposes making torture illegal, without exceptions, in order to raise the cost of torture sufficiently high so as to prevent the government from using torture unless it is absolutely imperative. In this case, the government officials should violate the law and hope for a pardon based on the merits of their decision. Similarly, Henry (1978) argues that even if torture might be necessary for national security purposes, it ought to remain illegal as to make this option least feasible by requiring one to defend his or her choice to torture legally and ex post. On the other had, economist James Miller makes another interesting argument for legalizing torture on convicted criminals on the grounds that torture, however unpopular, can be a more cost effective and no less immoral method of 4 Alex Tabarrok argument can be found at http://www.marginalrevolution.com/marginalrevolution/2005/11/torture_terrori.html. 6 7 criminal punishment than imprisonment.5 Miller argues that torture may not necessarily be any less immoral or criminal than prison rape and other abuses that many prisoners are subjected to as the result of their imprisonment. 5. Torture through the Lens of Conflict Literature Since torture is strongly linked to other types of human rights abuses, perhaps it should be examined in the broader context of violence and coercion. Torture appears to be highly correlated with extra judicial killings, disappearance of political opponents, political imprisonment as well as restrictions of basic civil freedoms such as freedom of speech, freedom of religion, freedom of movement, freedom of assembly and association, free and fair elections, and workers’ and women’s rights. Hence, a review of the conflict literature might be helpful in understanding why torture is so prevalent. Most of the economic analysis of conflict is based on the rational economic framework, which postulates that warfare becomes an attractive dispute resolution method when property rights or contract agreements are poorly defined and enforced because warfare becomes the enforcement mechanism in its own right. In his pioneering work, Schelling (1960) argues that nations with complete information should never go to war because a peaceful settlement is less costly than a conflict. However, the inability to enforce that peaceful settlement is what creates the need for war as a costly enforcement mechanism. By the same logic, torture is inferior to peaceful dispute resolution (i.e. interrogation) and should not occur unless its function is to be the enforcement mechanism. 5 James Miller’s argument for torture can be found here at http://www.overcomingbias.com/2007/08/bias-againstto.html. 7 8 However, even if torture is the necessary enforcement mechanism, its effectiveness in obtaining the desired level and accuracy of information shall not remain undisputed. According to Rejali (2004, 2007) and Koppl (2005), torture may not be any more effective in obtaining secret information than traditional human intelligence because of the fundamental problem of asymmetric information (Akerlof, 1970). For example, torture that yields inaccurate information may set anti-terrorist efforts on a wrong path and divert valuable resources away from preventing a massive terrorist attack as in the case of the ticking time bomb scenario. Koppl (2005) reasons that if the torturer knows nothing about the secret information set he is trying to extract, he is not able to verify whether the tortured subject is telling the truth or a lie. Without knowing the relevant information set, the torturer cannot vouch for the accuracy of the received information, but learning about this information set through torture is the objective in the first place. The torturer is trapped in a vicious circle due to the asymmetric information problem. Then, why are acts of torture so prevalent? Several attempts to reconcile these puzzles have been made. Perhaps, torture is not only used as a means of obtaining information but also as a means of maintaining social control and signaling a credible threat.6 An application of game theory with particular emphasis on reputation capital would be very useful in understanding why the use of torture is so persistent. The conflict literature might also provide some answers. For instance, Hirschleifer (1995) contends that anarchy is especially susceptible to conflict unless there are strong diminishing returns to fighting and incomes exceed the viability minimum. Fearon (1995) suggests that peaceful bargaining may not occur because of commitment problems due to incentives to renege on the bargain terms, issue indivisibilities (such as to legalize or not to legalize abortion), private 6 Wantchekon and Healy (1999) claim that “Torture can be a rational choice for both the endorsing state and the individual torturer” (p. 596) without sufficiently explaining why. 8 9 information about relative military capabilities or resolve, and incentives to misrepresent such information. Furthermore, Garfinkel and Skaperdas (2000) demonstrate that conflict could be the rational equilibrium outcome if the long-run gains from defeating an opponent outweigh the short-run losses. Nafziger and Auvinen (2002), on the other hand, find that income inequality and pervasive rent-seeking by the ruling elites may lead to war and state violence. The logic for conflict provided by these studies can be extended to explain the persistence of torture. 6. The Determinants of Torture and Other Types of Violence Whatever the true roots of violence and torture are, several studies show that geographical, institutional, and economic factors affect conflict and might influence torture use as well. A growing list of studies finds that democracies may engage in different types of conflicts and utilize different military strategies compared to totalitarian states, which implies that democracies might differ from autocracies in torture practices as well. Filson and Werner (2004), Garfinkel (1994), and Mitchell et al. (1999) argue that democratic regimes tend to accept negotiated settlements over wars and choose to fight only low cost, short wars that they can win. Persson (2002) and Mulligan et al. (2004) find that democracies differ significantly from autocracies when it comes to military spending, torture, execution, censorship, and religious regulation. Mulligan and Tsui (2006) argue that democracies and non-democracies should have identical policies except for cases when these policies influence the threat of entry from political challengers. In support of Mulligan and Tsui’s argument, Goldsmith (2003) and Yakovlev (2007) find that democracies spend proportionally less on national defense compared to dictatorships, holding everything else constant. 9 10 However, lower defense burdens as percentage of GDP do not turn democracies into weak military opponents. On the contrary, Biddle and Long (2004) argue that factors like superior human capital, harmonious civil-military relations, and Western cultural background are highly correlated with democracies and are largely responsible for democracy's apparent military effectiveness. Yakovlev (2008) finds that democratic and economic freedoms, volunteer armies, human capital, and GDP per capita increase country’s military capital intensity, which lowers conflict casualties. Horowitz et al. (2006) also find that democracies mitigate some of the conscription-induced casualties and are able to sustain high levels of casualties when targeted by authoritarian states. These findings suggest that torture and other forms of violence could have the same determinants. A growing body of interdisciplinary research on the determinants of torture and other human rights abuses has been emerging in the last two decades. Factors such as democracy, economic development, population size, wars, and acts of terrorism are typically found to have a significant effect on human rights violations (Bohara et al. 2006; Poe and Tate 1994; Davenport 1995, 1996; Poe, Tate, and Keith 1999; Richards, Gelleny, and Sacko 2001). Using district-level data on Maoist insurgency in Nepal, Bohara et al. (2006) find that an exchange of violence between government and opposition forces depends on the political and geographical factors. More specifically, their findings indicate that democracy and social capital have an inhibiting effect on the level of violence between government and opposition. In a cross-sectional timeseries analysis of 137 countries from 1950 to 1982, Davenport (1999) finds that autocracies appear to escalate and democracies tend to pacify state repressions. Henderson (1991) finds that democracy, socioeconomic conditions, inequality, economic growth, and economic development have a significant effect on the level of political repression. Democratic governments, as is 10 11 commonly argued, tend to limit the use of violence against their own people (Henderson 1991, 1993; Davenport 1995, 2004; Davenport and Armstrong 2004; Poe and Tate 1994; Richards 1999; Zanger 2000). Supposedly, democracies prefer to resolve conflicts through compromise, dialogue, political participation, and decentralized decisions rather than violence (Gurr 1986; Poe 2004; Rummel 1997). However, some studies find a non-linear relationship between government repression and political freedoms suggesting that most human rights violations occur in quasi-democratic countries (Fein 1995; Regan and Henderson 2002). Several studies indicate that even democratic countries can resort to repressions and torture when suffering from terrorist attacks, interstate conflicts, and civil wars (Krain 1997; Poe 2004; Poe and Tate 1994; Poe, Tate, and Keith 1999; Rasler 1986; Sherborne 2003; Zanger 2000). For instance, Dreher et. al (2007) discover that terrorist acts increase the probability of extrajudicial killings, political imprisonment, and torture in countries where they are committed, but terror has no effect on political participation, freedom of religion, freedom of speech, or freedom of movement. Hooks and Clayton (2005) argue that torture is not unique to autocracies as exemplified by systematical violence and prisoner abuse cases in supposedly democratic countries. According to Rejali (2007), most clean tortures (tortures that leave no permanent scars and therefore are hard to prove) were actually pioneered by imperial democracies such as Britain and France. Therefore, human rights groups may never be able to eradicate torture even in democratic societies because torturers resort to techniques that do not leave scars, like water boarding. Rejali (2007) concludes that most modern torture practices have strong historical path dependence and are passed on like a skill or craftsmanship. Economic development is also found to have a strong connection to torture (Carey 2004; Davenport 1995; Mitchell and McCormick 1988; Poe and Tate 1994; Poe, Tate, and Keith 1999). 11 12 According to Easterly et al. (2006), economic development leads to increased education, greater tolerance of other groups, higher productivity of labor and the value of human life. However, economic development could also make murderous political leaders more productive at killing since economic development brings with it advances in technology and social organization (Easterly et al. 2006). Abundance in natural resources may be less of a blessing and more of a curse when intense competition for resources in ethnically fragmented countries leads to ethnic cleansing, torture, and other human rights abuses. A pair-wise, cross-country correlation analysis in Table 1 reveals that the occurrence of torture decreases with increases in democracy, education, per capita income, military spending per soldier, and economic freedom index. On the other hand, torture appears to increase with an increase in energy consumption, total and urban population, government size, geographic elevation, and percentage of population living in the tropics. It could be argued that some of the above mentioned factors proxy for the level of economic development, resource scarcity, lack of government transparency, and ethnic fractionalization that make violence, in general, and torture, in particular, more appealing. [Table 1 about here] The pair-wise correlations in Table 1 are consistent with the previous findings on the importance of terrain (Fearon and Laitin 2003), democracy, and social capital (Weingast 1997; Putnam 1993; Russett and Oneal 2001) in affecting torture and violence. 7. On the Economic Theory of Torture (or Lack Thereof) 12 13 The literature on torture and conflict explored in the previous sections of this chapter raises a number of questions that warrant further empirical and theoretical research. For instance, why is torture so abundant despite its costliness and dubious effectiveness in gathering reliable information? Why does the treatment of POWs from different countries vary so much within one country (i.e. the case of Nazi Germany)? How can one ensure that human rights treaties are binding? Attempts by economists to answer these questions would be very valuable. The theoretical model developed by Wantchekon and Healy (1999) is a step in the right direction. They analyze the prevalence of torture in a dynamic game with incomplete information involving the state, the torturer, and the victim. The results show that when the state endorses torture as a means of extracting information or intimidating political opposition, torture is carried out with positive probability and becomes more widespread and cruel. Wantchekon and Healy (1999) claim that the information-gathering use of torture can be reduced by increase an individual’s ability to resist torture, perhaps, through a culture of solidarity and civil disobedience. When torture is used as a method of social control, only a revolt or international pressure can reduce torture. Otherwise, very few economists have devoted sufficient efforts to this topic, especially when it comes to theorizing about torture in the context of rational and self-interested economic agents. In this section, I propose a simple model of torture and derive the optimal level of torture that will be chosen. I also suggest potential avenues for future theoretical research on torture. Despite the overwhelming arguments made by economists in favor of free trade, antitrade policies remain to be widespread and hard to eradicate. The conventional explanation for the persistence of anti-trade policies despite their negative welfare impacts is that the 13 14 concentrated costs and widespread benefits of trade liberalization enable successful anti-trade lobbying. A similar argument can be made about torture. A soldier who tortures a prisoner of war for information gathering or other purposes opens the door to potential retaliation from the opposing side in the conflict (assuming perfect information). However, the same soldier is unlikely to be captured and administered the same punishment in retaliation to his decision to torture. Hence, the cost of potential retaliation from the enemy is spread across all soldiers who might be captured and tortured in return. The torturer captures full benefits from torturing a POW, but bears, at best, only a fraction of the costs from potential retaliation. If the person administering the torturous acts will never see a front line, then the retaliatory cost of torture for this individual is probably zero. The above argument is similar to the externality problem in economics where individual agents do not bear the full benefit or cost of their activity that generates that externality. The following game theoretic exposition explores this tendency in more detail. Assume that the two opposing forces are made up of homogenous agents facing an identical level of torture or ability to torture t, weighted by an identical positive probability of being captured c. Now, imagine a hypothetical soldier trying to decide whether or not to torture a captured enemy soldier for the purpose of extracting valuable information that could save his life or the lives of his comrades. A soldier weighs the payoffs from torturing versus not torturing and chooses the highest payoff strategy. Suppose the payoff for not torturing is some positive value q when there is no retaliatory torture from the opposing side of the conflict either. If the opposing side in this conflict chooses to torture, the payoff for the torturer is q+t and the payoff for the non-torturing enemy soldier is q-ct. If both sides of the conflict decide to torture, a soldier’s payoff on either side becomes q+t-ct or q+t(1-c). Figure 1 illustrates this payoff matrix for a soldier from side 14 15 one and a soldier from side two of this hypothetical conflict. From these payoffs, which are reminiscent of the prisoner’s dilemma game, it is obvious that the dominant strategy for both soldiers is to torture. [Figure 1 about here] Moreover, for values of c and t such that 0 < c < 1 and 0 < t < ∞, the payoffs in Figure 1 indicate that the strategy to torture is not only dominant but also socially optimal. This is where my analysis departs from the conventional prisoner’s dilemma outcome. Recall that in the prisoner’s dilemma game the socially optimal outcome is for both prisoners not to betray each other, but the individually optimal outcome is to betray each other, which is the correct solution. Both prisoners would be better off (i.e. it would be socially optimal) if they could ensure that neither betrays the other, but because they cannot enforce this agreement, the outcome of the game is individually optimal but socially suboptimal. However, my model’s solution differs from the prisoner’s dilemma because the payoff to torturer and the cost of being tortured are equal, which eliminates the tradeoff between what is socially and individually optimal for the two soldiers involved. This leads to a rather dismal conclusion that the decision to torture is both individually and socially optimal and will be chosen with a positive probability. If torturer’s payoff is lower than the cost to him of being tortured, the solution would be equivalent to the outcome in prisoner’s dilemma game. To describe the optimal level of torture to be chosen by each side of the conflict, let a soldier from side one (Soldier 1) maximize his utility by choosing the optimal level of torture 15 16 that is constrained by the probability weighted retaliation from side two soldier (Soldier 2).7 The objective function for Soldier 1 can be stated as follows: Max U1 = u1(t1) such that c1t2(t1) ≥ 0 (1) Let µ to be the Kuhn-Tucker multiplier on the constraint so that the Lagrangian takes the following form: £ = u1(t1) - µc1t2(t1). (2) The first order condition yields: ∂u1 ∂t = μc1 2 . ∂t1 ∂t1 (3) Soldier 2 solves a similar problem: Max U2 = u2(t2) such that c2t1(t2) ≥ 0. (4) Let µ to be the Kuhn-Tucker multiplier on the constraint. The second Lagrangian takes the following form: 7 Assume a utility function that is concave in torture and that 0 ≤ c ≤ 1 and 0 ≤ t ≤ ∞. 16 17 £ = u2(t2) - µc2t1(t2). (5) The first order condition for Soldier 2 is: ∂u2 ∂t = μc2 1 . ∂t2 ∂t2 (6) When µ, c1, c2, t1, t2 > 0, the first order conditions suggest that the optimal level of torture will be reached when the marginal benefit of torture torture retaliatory response μci ∂t j ∂ti ∂ui equals to the probability weighted, marginal ∂ti , where i = Soldier 1 and j = Soldier 2. Now, one can evaluate how changes in probability c can affect the level of torture used. As shown in Figure 2, a lower probability of being captured alive such as c* < c** would be associated with a higher level of torture such as t** > t* because of the diminishing marginal utility of torture (concave utility function). [Figure 2 about here] Figure 2 implies that a higher probability of being tortured by the opposing side reduces the level of desired torture. Alluding to the free trade analogy, a higher ci would imply a more equal distribution of costs and benefits from trade openness. Similarly, an increase in the marginal 17 18 retaliatory response torture ∂t j ∂ti by the opposing force j would also increase the marginal utility of ∂ui for side i, which means a lower level of torture due to diminishing marginal utility. ∂ti The above model can be made more intricate by incorporating asymmetric information about how, if at all, soldiers might be tortured and the importance of reputation capital in preventing the convergence of dominant strategy towards torture. This could be one of the avenues for future research. Since the role of government was omitted from this simple model, future studies could examine the effect of political regimes on the optimal level of torture. However, I suspect that the principal-agent problem would become a major obstacle in enforcing either domestic anti-torture policies or international treaties. Researchers may also want to examine how variations in the value of a statistical life across countries may affect their torture practices. For instance, if Soviet soldiers were relatively abundant compared to German soldiers in WWII, would that imply that they were proportionately less valuable than their German counterparts when it came to prisoner exchange and treatment of POWs? If so, this could explain the difference in treatment of the Soviet and American POWs in German prison camps. 8. Conclusion This chapter presents an introduction to torture from the rational economic perspective. A multidisciplinary literature review reveals that torture continues to be a widely used method of information gathering, coercion, and intimidation despite its questionable effectiveness and numerous treaties that prohibit it. Various economic, geographic, political, and institutional factors are found to be significantly related to torture. However, some of these factors have rather intricate relationships with torture. Democracy, for instance, may have a non-linear effect 18 19 on torture, which means that quasi-democratic countries may torture the most. Despite a wealth of empirical findings on torture, there is a lack of theoretical models of torture that rely on the rational agent framework. This chapter presents a simple model of torture between two warring countries based on the premise that torture costs are widespread and benefits are concentrated. The model’s solution indicates that torture will be carried out with positive probability. The optimal level of torture depends on the probability weighted level of retaliation by the opposing side. 19 20 References Akerlof, G. A. (1970), ‘The Market for ‘Lemons’: Quality Uncertainty and the Market Mechanism’, Quarterly Journal of Economics, 84(3): 488-500, 1970. Barro, R. J. and Jong-Wha Lee (2000), ‘International Data on Educational Attainment: Updates and Implications’, CID Working Paper No. 42. Available at http://www.cid.harvard.edu/cidwp/042.htm. Beevor, Anthony (2001), Stalingrad. Penguin Press. Bohara, A. K., N. J. Mitchell, and M. Nepal (2006), ‘Opportunity, Democracy, and the Exchange of Political Violence: A Subnational Analysis of Conflict in Nepal’, Journal of Conflict Resolution, 50: 108-128. Carey, S. C. (2004), ‘Domestic threat and repression: An analysis of state responses to different forms of dissent’, In Understanding human rights violations, edited by S. C. Carey and S. C. Poe, 202-20. Aldershot, UK: Ashgate. Cingranelli, D. L. and D. L. Richards (2008), ‘The Cingranelli-Richards (CIRI) Human Rights Dataset’, Version 2008.03.12. Available at http://www.humanrightsdata.org. 20 21 Davenport, C. (1995), ‘Multidimensional threat perception and state repression: An inquiry into why states apply negative sanctions’, American Journal of Political Science, 39 (3): 683-713. ——— (1996), ‘The weight of the past: Exploring the lagged determinants of political repression’, Political Research Quarterly, 49 (2): 377-405. ——— (1999), ‘Human Rights and the Democratic Proposition’, Journal of Conflict Resolution, 43: 92-116. Davenport, C. and D. A. Armstrong II (2004), ‘Democracy and the violation of human rights: A statistical analysis from 1976 to 1996’, American Journal of Political Science, 48 (3): 538-54. Dreher, A., M. Gassebner, and L. R. Siemers (2007), ‘Does Terror Threaten Human Rights? Evidence from Panel Data’, CESifo Working Paper Series No. 1935. Available at http://ssrn.com/abstract=971454. Easterly, W., R. Gatti, and S. Kurlat (2006), ‘Development, democracy, and mass killings’, Working Paper, Center for Global Development. Evans, Malcolm D. and Rodney Morgan (1998), Preventing torture: a study of the European Convention for the Prevention of Torture and Inhuman or Degrading Treatment or Punishment, Clarendon Press, Oxford University Press. 21 22 Fearon, J. D. (1995), ‘Rationalist Explanations for War’, International Organization, 49: 379414. Fein, H. (1995), ‘More murder in the middle: Life-integrity violations and democracy in the world, 1987’, Human Rights Quarterly, 17(1): 170-91. Filson, D. and S. Werner (2004), ‘Bargaining and Fighting: The Impact of Regime Type on War Onset, Duration, and Outcomes’, American Journal of Political Science, 48: 296-313. Fujiwara, Akira (1995), Nitchû Sensô ni Okeru Horyo Gyakusatsu, Kikan Sensô Sekinin Kenkyû, 9: 22. Gallup, J. L., J. D. Sachs and A. Mellinger (1999) ‘Geography and Economic Development’, CID Working Paper No. 1. Available at http://www2.cid.harvard.edu/ciddata/geographydata.htm. Garfinkel, M. R. (1994), ‘Domestic Politics and International Conflict’, American Economic Review, 84: 1294-1309. Garfinkel, M. R. and S. Skaperdas (2000), ‘Conflict Without Misperception or Incomplete Information: How the Future Matters’, Journal of Conflict Resolution, 44: 793-807. 22 23 Goldsmith, B. E. (2003), ‘Bearing the defense burden, 1886-1989: why spend more?’ Journal of Conflict Resolution, 47: 551-573. Gourevitch, Philip and Errol Morris (2008), Standard Operating Procedure, 1 edition, Penguin Press HC. Gurr, T. R. (1986), ‘The political origins of state violence and terror: A theoretical analysis’, In Government violence and repression: An agenda for research, edited by M. Stohl and G. A. Lopez, 45-72. Westport, CT: Greenwood. Gwartney, James and Robert Lawson (2004), Economic Freedom of the World: 2004 Annual Report, Vancouver: The Fraser Institute. Data retrieved from www.freetheworld.com. Hafner-Burton, E. M. and K. Tsutsui (2007), ‘Justice Lost! The Failure of International Human Rights Law To Matter Where Needed Most’, Journal of Peace Research, 44(4): 407-425. Hathaway, O. A. (2002), ‘Do Human Rights Treaties Make a Difference?’, The Yale Law Journal, 111(8): 1935–2042. ——— (2003), ‘The Cost of Commitment’, Stanford Law Review, 55(5): 1821– 62. ——— (2007), ‘Why Do Countries Commit to Human Rights Treaties?’ Journal of Conflict Resolution, 51: 588– 621. 23 24 Henderson, C. W. (1991), ‘Conditions affecting the use of political repression’, Journal of Conflict Resolution, 35(1): 120-42. ——— (1993), ‘Population pressures and political repression’, Social Science Quarterly, 74(2): 322-33. Heston, A., R. Summers and B. Aten (2002), Penn World Table, Version 6.1, Center for International Comparisons at the University of Pennsylvania (CICUP). Available at http://pwt.econ.upenn.edu/. Hirshleifer, J. (1995), ‘Theorizing About Conflict’, In Handbook of Defense Economics, Vol.1, edited by Keith Hartley and Todd Sandler, pp. 165-92. Amsterdam: North-Holland. Hooks, G. and M. Clayton (2005), ‘Outrages Against Personal Dignity: Rationalizing Abuse and Torture in the War on Terror’, Social Forces, 83(4): 1627-1645. Horowitz, M., E. Simpson and A. Stam (2006), ‘Domestic Institutions and Wartime Casualties’, Unpublished Draft Working Paper. Harvard University. Koppl, R. (2005), ‘Epistemic Systems’, Game Theory and Information 0510001, EconWPA. 24 25 Krain, M. (1997), ‘State-sponsored mass murder: The onset and severity of genocides and politicides’, Journal of Conflict Resolution, 41(3): 331-60. McCoy, K. E. (2005), ‘Trained to Torture? The Human Rights Effects of Military Training at the School of the Americas’, Latin American Perspectives, 32(6): 47-64. Mitchell, N. J., and J. M. McCormick (1988), ‘Economic and political explanations of human rights violations’, World Politics, 40(4): 476-98. Mitchell, S. M., S. Gates, and H. Hegre (1999), ‘Evolution in Democracy–War Dynamics’, Journal of Conflict Resolution, 43: 771-792. Mulligan, C. B., R. Gil, and X. Sala-i-Martin (2004), ‘Do Democracies Have Different Public Policies than Nondemocracies?’ Journal of Economic Perspectives, 18: 51-74. Mulligan, C. B. and K. K. Tsui (2006), ‘Political Competitiveness’, NBER Working Papers 12653, National Bureau of Economic Research, Inc. Nafziger, E.W. and J. Auvinen (2002), ‘Economic Development, Inequality, War and State Violence’, World Development, 30, pp. 153-163. European Union, Working Paper. Neumayer, E. (2005), ‘Do International Human Rights Treaties Improve Respect for Human Rights?’ Journal of Conflict Resolution, 49: 925-953. 25 26 North, Jonathan. 2006. Soviet Prisoners of War: Forgotten Nazi Victims of World War II, World War II . Persson, T. (2002), ‘Do Political Institutions Shape Economic Policy?’ Econometrica, 70: 883905. Poe, S. C., C. N. Tate and L. C. Keith (1999), ‘Repression of the human right to personal integrity revisited: A global cross-national study covering the years 1976-1993’, International Studies Quarterly, 43(2): 291-313. Poe, S. C. (2004), ‘The decision to repress: An integrative theoretical approach to the research on human rights and repression’, In Understanding human rights violations, edited by S. C. Carey and S. C. Poe, 16-38. Aldershot, UK: Ashgate. Polity IV Project (2000), ‘Political Regime Characteristics and Transition, 1800–2000’, Electronic data. (version p4v2000). College Park, Md.: CIDCM, University of Maryland. Rasler, K. (1986), ‘War, accommodation, and violence in the United States, 1890-1970’, American Political Science Review, 80(3): 921-45. Regan, P. and E. Henderson (2002), “Democracy, threats and political repression in developing countries: Are democracies internally less violent?” Third World Quarterly, 23(1): 119-36. 26 27 Reid, Patrick (1953), The Colditz Story, Greenwood Pub Group. Rejali, Darius (2004), ‘Does Torture Work?’ Salon.com, 21. Rejali, Darius (2007), ‘Torture and Democracy’, Princeton University Press. Richards, D. L. (1999), ‘Perilous proxy: Human rights and the presence of national elections’, Social Science Quarterly, 80(4): 648-65. Richards, D. L., R. D. Gelleny and D. H. Sacko (2001), ‘Money with a mean streak? Foreign economic penetration and government respect for human rights in developing countries’, International Studies Quarterly, 45(2): 219-39. Rolf, David (1998), Prisoners of the Reich: Germany's Captives, 1939-1945, Pen & Sword Books. Rummel, Rudolph (1997), Power Kills: Democracy as a Method of Nonviolence, New Brunswick, NJ: Transaction Publishers. Schelling, Thomas (1960), The Strategy of Conflict, Cambridge, MA: Harvard University Press. 27 28 Sherborne, L. (2003), ‘An integrated model of political repression: Theory and model development’, Ph.D. dissertation, University of Houston, Houston. Singer, J. D., S. Bremer and J. Stuckey (1972), ‘Capability Distribution, Uncertainty, and Major Power War, 1820-1965’, in Bruce Russett (ed) Peace, War, and Numbers, Beverly Hills: Sage, 19-48. Vreeland, J. R. (2008), ‘Political Institutions and Human Rights: Why Dictatorships Enter into the United Nations Convention Against Torture’, International Organization, 62: 65–101. Wantchekon, L. and A. Healy (1999), ‘The ‘Game’ of Torture’, The Journal of Conflict Resolution, 43(5): 596-609. Yakovlev, P.A. (2007), ‘Arms Trade, Military Spending, and Economic Growth’, Defense and Peace Economics, 18(4): 317–338. Yakovlev, P. A. (2008), ‘Saving Lives in Armed Conflicts: What Factors Matter?’ Economics of Peace and Security, 3(2): 68-73. Zanger, S. C. (2000), ‘A global analysis of the effect of political regime changes on life integrity violations, 1977-1993’, Journal of Peace Research, 37(2): 213-33. 28 29 Table. 1 Selected Pair-Wise Correlates for Torture (1) Democracy (2) Energy consumption per capita (3) Total population (3) Urban population (3) Real GDP per capita (4) Government consumption, % of GDP (4) Military spending per soldier (3) Economic freedom index (5) Average years of schooling (6) Elevation (7) Percentage of population in tropical area (7) -0.3077* 0.0539* 0.1819* 0.1817* -0.4871* 0.1931* -0.3563* -0.3908* -0.5415* 0.3583* 0.2865* Notes: * Statistically significant at the 5% level. The sample is an unbalanced panel of 200 countries from 1981 to 2006. Data sources: (1) Cingranelli and Richards, 2008; (2) Polity IV Project, 2000; (3) Singer et al., 1972; (4) Heston et al., 2002; (5) Gwartney and Lawson, 2004; (6) Barro and Lee, 2000; (7) Gallup et al., 1999. 29 30 Figure. 1 Prisoner’s Dilemma Applied to Torture Soldier 2 No torture Torture No torture q1, q2 q1-ct, q2+t Torture q1+t, q2-ct q1+t(1-c), q2+t(1-c) Soldier 1 30 31 Figure. 2 The Optimal Choice of Torture under Different Probabilities ui ti* ≈ ci** ti**≈ ci* ti 31