«© Nir Gazit and Robert J. Brym (2010) Department of Sociology University of Toronto 7,500 words Last revision: 26 April 2010 To be published in ...»
State-Directed Political Assassination in Israel: A Political Hypothesis*
© Nir Gazit and Robert J. Brym (2010)
Department of Sociology
University of Toronto
Last revision: 26 April 2010
To be published in International Sociology 26: 2011
* We thank Eitan Alimi, Bader Araj, Nava Löwenheim, Jack Veugelers, and two anonymous
reviewers for helpful comments on a draft. This paper is based on a project funded by the Social
Sciences and Humanities Research Council of Canada (File No. 410-2005-0026). Please direct correspondence to Robert J. Brym, Department of Sociology, University of Toronto, 725 Spadina Avenue, Toronto M5S 2J4, Canada (e-mail firstname.lastname@example.org).
State-Directed Political Assassination in Israel: A Political Hypothesis Introduction Extant theories explain reasonably well why the Israeli state exercises a given level of violence against substate actors. Based on economic or sociological models of human action, these theories attribute the level of state violence, respectively, to the narrow cost-benefit calculations of state officials or the institutionally embedded norms that govern their deliberations. The strength of such theories notwithstanding, we argue that they fail to account for the willingness of Israeli officials to order the assassination of high-ranking political opponents during the second intifada, or Palestinian uprising against Israel. Our analysis of published sources concerning the assassination of Hamas leaders Ahmed Yassin and Ismail Abu Shanab, and interviews with 74 Israeli counterterrorist experts, suggest that the decision to engage in statedirected political assassination in the period 2000-05 was based less on narrow calculations and institutionally specific norms than on identifiable political contingencies. Specifically, the second intifada appears to have led many Israeli decision makers to favour creating chaos in the Palestinian political system, a goal that was well served by the policy of political assassination.
The policy’s effect was to forestall the founding of a viable, independent Palestinian state.
Israel’s Assassination Policy Two months after the outbreak of the second intifada in September 2000, Israel openly acknowledged its policy of assassinating selected Palestinian insurgents (Abdel-Jawad, 2001;
David, 2002; Gross, 2003; Kasher and Yadlin, 2005). The United States initially expressed strong disapproval but muted its criticism following al-Qaeda’s attacks on the World Trade Center and the Pentagon in 2001. It all but fell silent on the subject after November 2002, when CIA operatives in Djibouti directed a Predator drone to fire a missile that killed a suspected alQaeda operative and planner of the 2000 USS Cole bombing. Insurgents who launch suicide attacks, lack easily identifiable bases of operation, and blend into local populations make it difficult to use large-scale conventional military force and increase the temptation to respond by unconventional means. The extraordinary nature of the violence visited on Israel and the United States has encouraged both countries to respond extraordinarily.
Since World War II, a new international norm has crystallized; either in court or on the battlefield, political leaders must answer for their states’ actions. Unusually, however, large segments of the public in Israel and the United States agree that this principle should extend to the assassination of political leaders they identify as terrorists. Public opinion polls show that as early as 1998 a majority of Americans favoured the assassination of individual terrorist leaders (Appleton, 2000: 507). Some 90 percent of Israelis support the policy (David, 2002: 7).
Accordingly, between October 2000 and July 2005, some 210 state-directed assassination attempts took place in Israel, the West Bank, and Gaza, resulting in 399 Palestinian deaths (Brym and Araj, 2006a).
Some students of Israel’s assassination policy focus on its legal and moral implications.
Such issues do not concern us here (David, 2003; Eichensehr, 2005; Kasher and Yadlin, 2005;
Keller and Forowicz, 2008; Kremnitzer, 2006; Kretzmer, 2005; Stein 2003). Instead, we are among those researchers who seek to explain variation in the level of Israeli retaliation against Palestinian insurgents, the consequences of such retaliation, and the logic (or lack thereof) underlying Israel’s counterinsurgency policy (Brym and Maoz-Shai, 2009; Byman, 2006; Cohen, 2008; Hafez and Hatfield, 2006; Honig, 2007; Kaplan, Mintz, Mishal and Samban, 2005;
Kaplan, Mintz and Mishal, 2006; Kober, 2007; Luft, 2003; Maoz, 2007; Mitchell, 2004; Ron, 2003; Zussman and Zussman, 2006). Specifically, this paper is motivated by an our finding that extant explanations account less well for variation in the rate of assassination of political leaders (about 9 percent of targets between October 2000 and July 2005) than for variation in the rate of assassination of military operatives (about 91 percent of targets during this period) (calculated from Brym and Araj, 2006a).1 We regard this shortcoming as serious because Israel’s assassination of political leaders typically causes more Palestinian outrage and intransigence, and has more negative implications for Israel domestically and internationally, than does the assassination of military operatives.
There were “only” 11 cases in the population of assassinated Palestinian political leaders between 2000 and 2005. We are therefore obliged to generalize cautiously, stating our findings An independent tally of type of assassination target for the period September 2000 to April 2004 is virtually the same as ours (Kober, 2007).
not as firm conclusions but as hypotheses requiring study in other times and different contexts.
Furthermore, we recognize that several assassinated political leaders also played important military roles in their organizations, just as civilian leaders in sovereign countries are sometimes in charge of their state’s security agencies. What distinguishes the 11 men we studied is that they all served as members of the central leadership (9 cases) or regional leadership (2 cases) of their political organizations. All other assassination targets lacked these characteristics.
To make our case, we first highlight the limitations of rational choice and new institutionalist theories – the two most popular theories that have been used to explain Israeli violence against substate actors. We then analyze the circumstances surrounding the assassination of Ismail Abu Shanab (21 August 2003) and Ahmed Yassin (22 March 2004), two of Hamas’s most senior political leaders. Yassin and Abu Shanab represented opposite ideological poles in Hamas. Yassin was Hamas’s top political and spiritual leader. Strongly influenced by the religious and political ideas of the Muslim Brotherhood, Yassin was a hardliner and one of the most radical voices opposing compromise with Israel. Abu Shanab, although subject to Yassin’s leadership and committed to Hamas ideology, expressed Hamas’s more moderate and pragmatic side. He supported the idea of a two-state solution and a long-term ceasefire with Israel. Consequently, analyzing the assassination of these figures ought to maximize our understanding of the broad strategic rationale underlying the Israeli policy of political assassination.
Our information comes in part from Hebrew-language newspaper accounts of, and published research reports pertaining to, state-directed political assassination during the second intifada, the Yassin and Shanab assassinations in particular. In addition, we analyzed a series of semistructured, approximately 90-minute interviews that were conducted in Hebrew in 2005 and 2006 by Yael Maoz-Shai with 74 senior Israeli security decision makers and advisers from various state and academic organizations (for sampling details, see Brym and Maoz-Shai, 2009).
Israeli decision makers regard Palestinian political leaders as terrorists, and thus as legitimate targets for assassination. However, the evidence we present is consistent with the view that the strategic motivation underlying the policy of assassinating political leaders is broader than the desire to thwart terrorism. It seems to encompass the wish to promote instability in the Palestinian polity, thus delaying if not preventing the creation of a viable Palestinian state.
Extant Explanations and their Shortcomings Researchers who endorse an economic model of human action have demonstrated that the Israeli state typically behaves rationally in a narrow sense: the level of violence wielded by the state is strongly and positively correlated with the severity of preceding attacks on its citizens, while the exercise of state violence typically brings about a decline in anti-Israel attacks in the short term (Almog, 2004-05; Berrebi and Klor, 2006; Bronner, 2009; Frisch, 2006; Jaeger and Paserman, 2006; 2008; Morag, 2004).2 Generally, rational choice theorists do not concern themselves with the social origins of the strategies and goals that frame human action because they assume that particular means and ends are optimal and therefore “given” (Kiser and Bauldry, 2005: 173). In the case of the Palestinian-Israeli conflict, they usually fail to discuss the rationality of responses to anti-Israel violence other than measured state violence, and outcomes other than a short-term decline in violent action on the part of substate actors, because they apparently assume that alternative means and outcomes are suboptimal and therefore nonrational.3 Researchers who favour a sociological model of human action take a different tack. They argue that norms associated with different institutional settings affect the level of violence visited on substate actors, to some degree independently of the strategies and goals of state officials.
Researchers in this tradition have little if anything to say about long-term effects, but at least one examination of the use of limited force against substate actors from 1949 to 2006 discovered that, in the long term, aggressive actions by Israel consistently failed to lower the frequency and lethality of enemy attacks and had adverse military and diplomatic effects (Maoz, 2007).
Classical rational choice theorists claim that “widespread and/or persistent human behavior can be explained by a generalized calculus of utility-maximizing behavior, without introducing the qualification ‘tastes remaining the same’” (Stigler and Becker, 1977: 76). For classical rational choice theory, tastes (or “preferences,” to use the now more common term) are similar among people and remain stable over time. Therefore, the explanation of any particular behavior requires only the discovery of the utility-maximizing principle(s) governing it, not the discovery of the origins of preferences, which one may safely assume to be fixed.
Following are three examples of new institutionalist arguments and findings relevant to the case
1. States are not free to use violence as their governments might like in response to threat from dissenting minorities because the international community imposes constraining norms on them. Defying international norms incurs costs, knowledge of which limits state action (Gordon and Berkovitch, 2007; Hajjar, 2001; 2005; Risse, Ropp, and Sikkink, 1999; Shor, 2008).
2. The stronger the economic ties and bureaucratic state regulation between core states and their occupied territories, the greater the degree to which core-state norms influence the treatment of dissenting minorities, and therefore the lower the likelihood of core states engaging in acts of indiscriminate violence against dissenting populations residing in occupied territories (Ron, 2000; 2003).
3. Organizationally embedded conventions guide the response of state decision makers to threat. These conventions remains intact until changing domestic and international circumstances cause growing awareness of their non-viability, at which time a new set of conventions is created. The very definition of what constitutes threat, the maximum level of violence by dissenting minorities that is deemed tolerable, the appropriate means for dealing with such violence, and the desirability of various outcomes that flow from such violence are thus shaped by institutionalized norms that vary over time (Kuperman, 2005).
The foregoing arguments are based on research findings, not speculation. Rational choice and new institutionalist theories offer credible, evidence-based accounts of variation in state violence against dissenting minorities in a variety of settings. However, saying that the theories have strengths does not mean that they lack flaws.
In the case of the Israeli-Palestinian conflict, we are struck by the strong correlation between the frequency of Israeli deaths at the hands of Palestinians in a given month and the frequency of Palestinian deaths at the hand of Israelis in the following month (r =.629 for the period January 1987 to December 2007). Also noteworthy is the more modest but still positive correlation between the frequency of Palestinian suicide bombings in a given month and the frequency of Israeli state-directed assassination of military operatives in the following month (r =.241 for the period October 2000 to July 2005). However, the correlation between the frequency of Palestinian suicide bombings in a given month and the frequency of Israeli statedirected assassination of Palestinian political leaders in the following month is close to zero (r =.097, again for the period October 2000 to July 2005).4 True to rational-choice theory, more Palestinian suicide bombings in one month tend to be followed by more Israeli killing of Palestinians, including more assassinations of military operatives. However, the argument that Israeli response is proportionate to Palestinian threat appears to break down when applied to the assassination of political leaders. In the latter case, the narrow game rules specified by adherents of rational choice theory seem not to apply.