Annexation Attempts as a Two-Level Game: Israel and the West Bank in 1967 and 2020

Abstract

Annexation attempts threaten international security and the rules-based world order. Yet, studies of annexation are rare, whilst the scant relevant literature is great-power centric. This article therefore asks why some non-great power annexations succeed, whereas others do not. Applying Putnam's two-level game framework, it analyses an occupier's: (1) domestic politics; (2) international relations; and (3) interactions between these two levels of analysis. It applies this framework to Israeli policy in two specific cases: partial annexation, where Israel annexed East Jerusalem but not the entire West Bank (1967); and Israel's comprehensive but aborted West Bank annexation (2020). This article finds that when the policy would yield maximum domestic returns and minimal global opprobrium, Israel's leaders enacted annexation. They refrained from doing so when this synchronization was absent. These findings illustrate the utility of the two-level game framework for explaining non-great powers’ decision-making and their territorial policies in particular.

Resumen

Los intentos de anexión amenazan la seguridad internacional y el orden mundial, el cual está basado en normas. Sin embargo, los estudios en materia de anexión son escasos, y la escasa literatura relevante existente tiende a centrarse en las grandes potencias. Por lo tanto, en este artículo nos preguntamos por qué algunas de las anexiones llevadas a cabo por países que no son grandes potencias tienen éxito, mientras que otras no. Aplicamos el marco de juego de dos niveles de Putnam con el fin de analizar lo siguiente desde el punto de vista de los ocupantes: i) la política interna, (ii) las relaciones internacionales, y (iii) las interacciones entre estos dos niveles de análisis. En el artículo aplicamos este marco de trabajo a la política israelí en dos casos concretos: la anexión parcial, que tuvo lugar cuando Israel se anexionó Jerusalén Este, pero no toda Cisjordania (1967), y la anexión integral, que fue abortada, de Cisjordania por parte de Israel (2020). Este artículo concluye que los líderes de Israel promulgaron la anexión cuando esta política conllevaría el máximo beneficio a nivel interno y el mínimo oprobio global. Por el contrario, se abstuvieron de hacerlo cuando esta sincronización estaba ausente. Nuestras conclusiones ilustran la utilidad que tiene el marco de juego de dos niveles para explicar la toma de decisiones por parte de los países que no son grandes potencias, así como sus políticas territoriales, en concreto.

Résumé

Les tentatives d'annexion menacent la sécurité internationale et l'ordre mondial basé sur des règles de droit. Pourtant, les études portant sur l'annexion se font rares, quand la maigre littérature pertinente est centrée sur les grandes puissances. Cet article cherche donc à répondre à la question suivante: pourquoi certaines annexions de puissances inférieures réussissent-elles, alors que d'autres échouent? En appliquant le cadre du jeu à deux niveaux de Robert Putnam, il analyse: (i) la politique nationale d'un occupant; (ii) ses relations internationales et; (iii) les interactions entre les deux niveaux d'analyse. Il applique ce cadre à la politique israélienne dans deux cas précis: l'annexion partielle, quand Israël a annexé l'est de Jérusalem, mais pas la totalité de la Cisjordanie (1967) et l'annexion totale, mais avortée, de la Cisjordanie par Israël (2020). Cet article remarque que lorsque la politique générait un maximum de retombées sur le plan national et un minimum d'opprobre au niveau mondial, les dirigeants israéliens optaient pour l'annexion. Ils s'en abstenaient quand cette synchronisation ne se produisait pas. Ces résultats illustrent l'utilité du cadre du jeu à deux niveaux quand il s'agit d'expliquer la prise de décisions des puissances de second ordre et, plus précisément, leurs politiques territoriales.

Annexation, Israel, West Bank, Territorial Conflict

Anexión, Israel, Cisjordania, Conflicto territorial

annexion, Israël, Cisjordanie, conflit territorial

Introduction

In contemporary geopolitics, an extensive set of norms forbids territorial expansion by force. The main legal construct that enforces this prohibition is the distinction between the temporary occupation of a foreign territory and its permanent annexation (Roberts 2006). The latter half of the twentieth century saw these legal and normative prohibitions diffuse worldwide. Domestic audiences became decreasingly tolerant of indefinite interventions, whilst occupied peoples steadfastly resisted would-be conquerors (Edelstein 2008). International organizations and statutes, from the UN Charter to the Geneva Convention, now proscribe annexation. In previous centuries, states regularly annexed occupied foreign territory, but the institutionalization of a “territorial integrity norm” reversed this trend (Korman 1996; Altman 2020). Annexation became practically more difficult and, normatively, less acceptable.

Yet, annexation never disappeared. In 1975, Morocco annexed the occupied Western Sahara. In 1981, Israel annexed the Golan Heights. The fact that the global hegemon—the United States—subsequently recognized these annexation attempts in 2020 and 2019, respectively, suggests the prohibitions against annexation are under renewed stress (Kattan 2019). This trend was also illustrated when Russia occupied and subsequently annexed swathes of eastern Ukraine in September 2022 (Sauer and Harding 2022). These developments have precipitated predictions that “attempts to seize small territories will shape the twenty-first century” (Altman 2020, p. 517). Though annexation is apparently more costly and less legitimate, a minority of states still annex foreign territory, disregarding global norms and legal prohibitions.

Relatedly, this article asks, What processes, inputs, and actors render a state likely to annex an occupied territory, or refrain from doing so? International relations scholars rarely focus on modern annexation attempts, even though annexation is “the most extreme form of expansionism a state can pursue” (Maass 2017, p. 26). Emerging work on great power annexation has partially mitigated this deficit. By contrast, non-great power annexation attempts remain a “ripe area for further research” (Maass 2017, p. 8).

This article addresses this research gap. The contemporary research on both non-great powers and territorial conflict largely concurs that it is interactions between domestic and international politics that determine a state's policies. By contrast, the great power-centric annexation literature suggests that domestic politics exerts the most agency. To test these competing assertions, this article draws on Putnam's “two-level game” (TLG), which scholars have employed to explain how internal and external bargaining shape numerous manifestations of state behavior. It has, however, never been applied to annexation. Accordingly, this article scrutinizes three levels of analysis and interaction types: (1) between actors operating within the occupier's domestic politics; (2) between the occupier and other actors in the international system; and (3) between the domestic and international levels of analysis.

This article applies the framework to two case studies: Israel's 1967 decision to annex East Jerusalem but not the remainder of the West Bank; and Israel's 2020 West Bank annexation plan. These cases differ in that one (1967) was covertly implemented after territorial conquest. The other (2020) was an overt annexation attempt after decades of occupation. They also provide two divergent outcomes: One successful but partial annexation (1967) and one comprehensive but abandoned annexation attempt (2020).

These cases also present a research puzzle. In 1967, a relatively dovish party led Israel's government. Ministers expressed fears that annexation would precipitate unprecedented international opprobrium. Nonetheless, Israel annexed parts of the West Bank. In 2020, a hawkish government won re-election after pledging a comprehensive West Bank annexation. The West Bank also contained more Israeli settlers than ever before, creating a burgeoning pro-annexation constituency. Yet, Israel did not annex any of the West Bank. This article thus asks two questions: why did Israel enact limited annexation in 1967 and no annexation whatsoever in 2020, despite the increasing salience of pro-annexation domestic inputs in the latter case? Why, in both cases, did Israel refrain from comprehensively annexing the entire West Bank?

This article demonstrates that in 1967 and 2020, Israel's leaders avoided a comprehensive annexation because they perceived that the policy's costs outweighed any benefits. As Putnam predicts, it was a complex bargaining game between the domestic and international levels of analysis that precipitated this perception. This commonality engendered two divergent outcomes: limited annexation (1967) and non-annexation (2020). The TLG demonstrates that these two outcomes were both utility-maximizing. In 1967, there existed significant pro-annexation domestic pressure, but Israelis were divided between backing a comprehensive or limited annexation. External inputs—primarily US pressure—resolutely rejected a comprehensive annexation but were more ambivalent about a limited annexation. Israel thus implemented partial annexation. In 2020, this outcome was suboptimal, because powerful domestic hardliners would only endorse comprehensive annexation. International inputs, however, offered significant benefits for non-annexation and substantial costs for any annexation. Israel, accordingly, abandoned its annexation attempt.

This article first clarifies understandings of annexation and annexation attempts before examining Israel's West Bank policies since 1967. It makes three key contributions. Firstly, it constructs and tests a framework for assessing what makes annexation attempts succeed or fail. Secondly, it illustrates what external and internal inputs precipitate or restrict non-great power annexation attempts. Finally, it demystifies key Israeli case studies. Overall, this article helps advance both case-specific and broader understandings of annexation, which remains a contemporary geopolitical security challenge.

Annexation Attempts in Theory and Practice

Defining Annexation Attempts

States become occupiers when they capture and control a territory over which they previously did not exercise de facto or de jure authority. International law forbids an occupier from drastically altering the status quo within that territory and demands the occupier not make its control permanent (Roberts 2006). Annexation is, comparatively, all-encompassing, since it reveals the occupier's intent to absorb the territory, excluding withdrawal from the list of legitimate policy options (Fazal 2008).

Annexation may appear a straightforward concept, but its application is often contested. In 1977, a journalist asked Israeli Prime Minister Menachem Begin if he would annex the West Bank. Begin, in turn, replied: “You annex foreign territories, not your own territories that are liberated” (Avner 2013). The UN mandates that member states “refrain [. . .] from the threat or use of force against the territorial integrity or political independence of any state,” thereby proscribing annexation whilst avoiding explicitly defining the term (United Nations Charter 1941). The International Committee of the Red Cross (ICRC) defines annexation as a “unilateral act” where a state “proclaims its sovereignty over the territory of another” (International Committee of the Red Cross N.D.). However, this excludes scenarios where a state may not proclaim annexation but furtively annexes territory. The ICRC definition also fails to delineate between successful annexations and “annexation attempts,” where an occupier makes “substantial effort” to annex a territory, but then fails to do so (Maass 2017, p. 7).

To capture the broad gamut of the concept, this article employs Diel and Goertz's definition of annexation as: “when one political entity unilaterally extends its sovereignty over a piece of territory” outside of its declared and recognized borders (Diehl and Goertz, 1992, p53). The unilateral aspect distinguishes between annexation—which international norms and law deem illegitimate—and territorial changes, which are permissible when conducted consensually between the relevant contracting parties (Kacowicz 1994).

This article focuses on actor behavior (attempting to annex) rather than an outcome (annexing territory) (Altman and Lee 2022). Modifying Diehl and Goertz's formula, annexation attempts constitute when one political entity unilaterally attempts to extend its sovereignty over a piece of territory outside its declared and recognized boundaries. This definition allows for the scrutiny of what alignment of variables was present in annexations, but absent in annexation attempts. Equally, it focuses on a state's end goal, rather than requiring that the occupier operationalize intent in a specific way, for instance, by publicly proclaiming annexation. It permits the comparative scrutiny of declarative annexation attempts and furtive annexations, the latter of which are increasingly common in global politics.

Annexation Attempts in the Literature

Why do annexation attempts persist, despite legal and normative prohibitions? Within the broader literature on territorial conflict, rationalists and particularly realists argue that acquiring territory represents “the supreme political objective in a world of territorial states” in an anarchic world (dis)order (Mearsheimer 2001, p. 86). Not all territory is equally valuable. States prioritize annexing territory with “tangible” value that increases their security and power. Tangible characteristics span a territory's economic, military and strategic worth from resources within it to defensive frontiers, such as mountains or rivers, and its ability to provide a physical buffer against rivals (Liberman 1996; Walter 2003; Carter and Goemans 2011). Subjectivists and constructivists, conversely, stress a territory's relational, “intangible” value. Territory possesses intangible value if it is linked to hegemonic values within a given state. Factors that can precipitate this perception include an historic link to the territory and the presence of cultural/religious sites. A state is more likely to annex that territory and frame it as indivisible because losing control over it is perceived in existential terms (Goddard 2006).

Both rationalist and subjective analyses illustrate that annexation attempts are a distinct form of territorial behavior. In comparison to temporary occupations, annexation offers the “deepest possible influence over external territory by absorbing it within the state itself” (Maass 2017, p. 5). To rationalists, annexing territory positively affects a state's position in the global balance of power, by providing it with more resources and strategic options (Liberman 1996; Huth 1998). Subjectivists illustrate that annexation can re-shape hegemonic values far more than short-term occupations, by enabling states to “redefine their national homeland, moulding local identities, institutions and cultural politics” (Maass 2020, p. 4). In sum, both unit- and system-level inputs drive annexation attempts, whilst annexation can significantly alter a state's internal and external balances of power.

Correspondingly, interactions between domestic and international politics shape a state's territorial policies. Zellman (2020) illustrates how communal values can either resolve or perpetuate territorial disputes. Bregman (2015) highlights how it is both interactions between the occupier and third parties and between occupier and occupied that determine whether an occupation ends or is perpetuated. These two “circles of occupation” are complex systems. It is, however, interactions between them that determine an occupier's policies. Kacowicz (1994) posits that a state cedes territory only if it achieves an optimal bargaining outcome between its internal and external political dynamics. Nonetheless, none of these studies scrutinize one specific form of state behavior in territorial conflict: annexation.

Annexation Attempts and Non-Great Powers

Studies of non-great powers share the above deduction that interactions between a state's internal and external spheres determine its foreign policy (David 1991; Barnett and Levy, 1991; Pedersen and Reykers 2023). Non-great powers have less resources, influence and agency in international affairs than great powers, rendering them relatively susceptible to external pressure (Alons 2007; Long 2022). Further, a shared proposition that non-great powers behave differently than great powers underlines the emerging research (Long 2022; Tinh and Ngan 2022; Pedersen and Reykers 2023). As such, generalizations garnered from the study of great powers—IR's traditional focus—are less applicable to non-great powers.

This perspective sits uncomfortably with recent studies of annexation. Altman and Lee find that it is domestic political dynamics that drive conquest attempts. Their dataset, however, does not distinguish between great powers and other states (Altman and Lee 2022). The only recently-published IR book explicitly analyzing annexation scrutinizes one milieu alone: the US—a great power—and argues that external interactions lack agency in annexation decisions (Maass 2021). It explicitly notes, though, that the US was unusually insulated from “external constraints,” even when compared to “most great powers in history” (Maass 2020, p. 9). Other authors affirm that great powers and non-great powers behave differently in territorial conflicts (Diehl and Goertz, 1992). But, when non-great powers do feature in the literature, they appear largely as passive victims of great power annexation attempts (Fazal 2008).

In contrast, Morocco's annexation of Western Sahara and Israel's annexation of the Golan Heights illustrate that annexation attempts are neither a purely historical phenomenon, nor restricted to great powers. Annexation is a constitutive feature of post-colonial state formation, because newly-formed states often annex foreign territory to consolidate their internal and external legitimacy. Given that these new states’ rivals are usually other post-colonial states it exemplifies that non-great powers are both responsible for multiple contemporary annexation attempts and are disproportionate victims of annexation (Fazal 2008). Hence, there exists a literature gap and significant contemporary relevance for studies of annexation attempts that scrutinize non-great powers.

Building a Framework for Understanding Non-Great Power Annexation

The Two-Level Game Framework

Whilst the literature on territorial conflict, annexation, and non-great powers possesses divergent foci and causal claims, they share several pertinent convergences. Firstly, annexation is a distinct phenomenon that requires its own analytical framework. Paradoxically, though, few studies distinguish between annexation and other state behaviors in territorial conflicts (Muenger 1991; Liberman 1996). Secondly, great powers and non-great powers behave differently. Yet, no existing studies test this claim in annexation attempts, whilst existing studies of this phenomenon are either great power-centric or do not account for the divergencies between these actor-types. Third, when seeking to understand a state's territorial disputes, “the domestic and international levels of analysis must be joined in a systemic way” (Huth 1998, p. 17). However, no such framework has been applied to annexation attempts.

This article addresses this literature gap, by applying Putnam's two-level game to annexation attempts. Writing in 1988, Putnam challenged the hegemonic emphasis on systemic inputs in states’ foreign policies. Instead, Putnam argued that a state's leaders constantly balance simultaneous pressure from international (level I) and domestic (level II) actors. International negotiations succeed when decision-makers secure a “win-set,” the alignment of variables where a state would endorse an agreement. Several factors determine the possibility of this scenario, or win-set “size.” For an outcome to constitute a win-set, its perceived benefits to the governing coalition and the state itself must outweigh the cumulative level I and level II costs. It must also secure a critical mass of domestic and international support. In short: states will only consent to a level I agreement if a governing party can induce a majority of level II actors to endorse that agreement (You 2016; Bjola and Manor 2018). The dynamics that shape win-set size and the potential to fulfill the above criteria span level I bargaining and negotiation strategies, to level II institutions and coalitions (Putnam 1988).

The diametrically opposed demands of actors in each level of analysis shrink win-set size since, “moves that are rational for a player at one board [. . .] may be impolitic for that same player at the other board” (Putnam 1988, p. 434). An inter-state agreement may enhance a state's level I position in the global balance of power. But, if enough level II actors oppose that deal, consenting to it would imperil a government's survival. Level I intransigence may harm a state's foreign relations, but if a critical mass of level II actors support that stance, it could enhance a government's legitimacy. In this scenario, the level II benefits of opposing an agreement would outweigh the level I costs. Accordingly, level I actors intervene in each other's domestic politics, to entice a political realignment or “reverberation” to mitigate level II resistance (Putnam 1988, p. 456).

Subsequent works further test the TLG's explanatory value, whilst expanding on the framework's causal paths. Whereas Putnam limited the TLG to formal level I summits, other scholars have illustrated its applicability to informal bargaining scenarios (Mo 1994; You 2016; Murphy 2017; Thu and Tinh, 2023). The literature has expanded its scrutiny of level II inputs to include domestic interest groups, public opinion, and hegemonic values (Mo 1995; Shamir and Shikaki 2005). Authors have applied the TLG to ethno-territorial disputes, including the Arab-Israel conflict (Gong and Choi, 2019; Shamir and Shikaki 2005; Ben-Porat 2006; Wenger and Chen 2017). Yet it has never been applied to annexation attempts or to scrutinize Israel's territorial policies.

Assumptions and Hypotheses

Applying the TLG framework to annexation attempts begets several assumptions. Putnam and subsequent authors assert that level I bargaining preferences a shared optimal outcome, whereas level II bargaining impedes this scenario. This is because level II actors seek to prioritize policies that preserve their interests, and at least some of these actors will perceive that a level I agreement threatens these interests. As such, level II actors are more likely to produce hard-line inputs, which make a level I agreement less likely (Putnam 1988; Iida 1993). Level II actors are particularly likely to produce anti-concessionary inputs when the policy under negotiation is deemed particularly contentious (Wenger and Chen 2017). In these scenarios, governments are under greater threat at level II, if they demonstrate moderation at level I (Hurst, 2016). Applying the TLG to territorial conflicts illustrates that level II intransigence often impedes “straightforward boundary negotiations” (Manicom, 2014, p. 165). This suggests that annexation attempts—which are inherently contentious in that they re-shape a state's size, political order, and values—will elicit level II intransigence, particularly when the territory possesses tangible and/or intangible value.

The formulae below apply Putnam's assumptions to unpack the multilevel interplay of inputs that makes annexation more or less likely. The greater the perceived costs across both levels, the less likely an occupier is to annex territory. If the policy's costs are greater than or equal to a comprehensive annexation's perceived benefits, then comprehensive annexation—when an occupier operationalizes the most hard-line level II territorial demands—does not constitute a win-set (Alignment I). If the policy's perceived benefits outweigh its Level I and Level II costs, however, comprehensive annexation does constitute a win-set (Alignment II).

L_I + L_II costs < L_I + L_II benefits = ↑ Annexation

Alignment I: The alignment of inputs that makes annexation more likely

L_I + L_II costs >= L_I + L_II benefits = ↓ Annexation

Alignment II: The alignment of inputs that makes annexation less likely

Further, this article assumes that though dynamics within global and domestic politics affect an occupier's calculations, it is the interaction between actors and processes across these levels which ultimately prevent or precipitate annexation. This bargaining process begins within the international system (level I) and the occupier's domestic political environment (level II) simultaneously. As figure 1 illustrates, these inputs coalesce at the decision-making level into: (3) bargaining between domestic and international politics. It is this third level that determines which of the three outcomes in figure 1 constitutes a win-set. The interaction between these levels is a reflective process: Changes in level I dynamics can cause level II “reverberations,” whilst level I actors tailor their demands based on what level II audiences will accept. Whilst actors attempt to influence others operating at their own level and beyond, governing coalitions synchronize their win-sets across level I and level II.

Figure 1.

How multilevel bargaining coalesces to shape policy.

Because Putnam's original TLG framework scrutinized international summits, it presumed a binary of outcomes: that the states involved reach an agreement (success) or do not (failure) (Putnam 1988). As figure 1 demonstrates, this article instead assumes three potential outcomes: comprehensive annexation, partial annexation, or non-annexation. This is because historical annexation attempts produced more diverse outcomes than the international summits Putnam examined. Some states abandoned their annexation attempts altogether. Others jettisoned comprehensive annexation to instead annex small parts of a territory; others still have caused “state death” by comprehensively annexing a rival's entire sovereign territory (Fazal 2008; Maass 2021). If comprehensive annexation matches the Alignment I formula above and non-annexation corresponds to Alignment II, partial annexation lies in between. For instance, a comprehensive annexation could incur more costs than benefits. Yet perpetuating the status quo could also entail unacceptable costs if there exist significant pro-annexation pressure. As a result, a compromise option—partial annexation—would best maximize a state's benefits whilst also keeping costs relatively low.

Regardless of an annexation attempt's eventual outcome, tangible and intangible factors within both level I and level II shape decision-making. These span: the territory's material and non-material worth; hegemonic level I values; hegemonic level II values; a state's level I and level II balance of power; and the impact of annexation at level I and level II. Putnam applied the TLG to scenarios where level I bargaining sought to change a status quo through a negotiated agreement, whilst mitigating level II intransigence. This article reverses this dynamic: whilst level II remains an assumed source of intransigence, it is these pro-annexation inputs that seek to elicit a territorial change, whilst level I inputs seek to preserve a status quo.

Case Selection

This article applies the TLG to two multiple-embedded cases: (1) Israel's 1967 decision to annex East Jerusalem but refrain from annexing most of the West Bank; and (2) Israel's 2020 comprehensive, but aborted, West Bank annexation attempt. A multiple-embedded methodology combines the depth of single case analysis with the comparative inference of multiple cases, by examining several instances of a phenomenon within the same milieu (Yin 2009). These cases span two of the outcomes from figure 1: non-annexation and partial annexation. Equally, they constitute two divergent annexation attempts: furtive, successful but partial annexation in 1967, when Israel applied sovereignty over East Jerusalem but obfuscated the decision; and proclaimed, comprehensive but aborted annexation in 2020, when Israel announced it would annex most of the West Bank, but then refrained from doing so (table 1).

Table 1.

this article's two case studies

Occupier	Territory	Year	Annexation Attempt	Implemented
Israel	West Bank	1967	Partial (East Jerusalem)	Yes
Israel	West Bank	2020	Comprehensive (most of the West Bank)	No

Further study of the Israel-West Bank relationship is exigent, given its central role in perpetuating the Israeli-Palestinian conflict and shaping level I bargaining (Waxman 2008; Bregman 2015). In 2002, every Arab and Muslim-majority nation offered to normalize relations with Israel in exchange for the latter's full withdrawal from the West Bank and other territories occupied in 1967. Yet in January 2024, Israeli Prime Minister Benjamin Netanyahu demanded indefinite Israeli security control over the entire West Bank and rejected any withdrawal (Beaumont 2024). Israel has even spurned salient pressure from its most prominent ally and the global hegemon—the US—which has repeatedly sought to engineer an Israeli West Bank exit (Raz 2012; Satloff 2021). This suggests that level I inputs have sought to induce withdrawal and stymie annexation but have faced significant impediments.

As the TLG framework assumes, the literature illustrates that it is Israel's level II inputs that have hindered level I bargaining, given the West Bank's intangible and tangible value. The territory hosts multiple Jewish religious and archaeological sites. Concurrently, one commentator argues that “almost anything Israel does can be sooner or later tied to security concerns over the West Bank” (Cohen 1985). Israeli planners have long seen the West Bank as essential for “strategic depth:” controlling territory to create a buffer between Israel and its rivals (Eiran 2019). Further, pro-annexation pressure has increased, the longer that Israel's occupation has continued. This is because successive Israeli governments have encouraged Jewish settlement in the West Bank. The number of Israeli Jews within the territory has soared to over 500,000, suggesting that ethnic kinship constitutes an increasingly powerful, pro-annexationist level II input (Zellman 2015; Eiran 2019).

Nonetheless, this article assumes that in the case of the West Bank, level II inputs will not exert pro-annexation pressure alone. A former Deputy Head of Israel's National Security Council claims: “Everyone in Israel would support annexation; security, historical rights, we have more than enough reasons. The problem is the Palestinians” (Yaar 2017). Comprehensively annexing the West Bank would require that Israel decide whether to incorporate the territory's Palestinian residents as full citizens. Doing so would engender Israel's political and demographic status as a Jewish-majority democracy. On the other hand, annexing the territory whilst depriving Palestinians of full citizenship would suggest that Israel would no longer constitute a democracy. Accordingly, Israel's leaders have repeatedly equivocated on annexation and deferred a final decision on the territory's fate to future administrations (Pinfold 2021).

Putnam creates a dichotomized ideal-type of level II actors: “isolationists”—who ignore level I inputs—and “internationalists,” who favor a level I agreement. This article modifies this heuristic to apply it to annexation and the Israeli milieu. It classifies relatively pro-annexationist Israelis as unilateralists. Like Putnam's isolationists, these actors are most immune to level I pressure and advocate that Israel unilaterally annex all or parts of the West Bank. Concurrently, this article mimics Putnam's internationalist classification, given these individuals are comparatively attentive to level I inputs. They are therefore most likely to resist annexation, given the international norms and laws that discourage it. Putnam theorized that isolationists are pro-status quo since they seek to scupper any potential level I agreement, whereas internationalists are not, given that they prefer a new level I deal. By contrast, this article assumes that it is unilateralists who seek to alter Israel's territorial status quo, whereas internationalists are more wary of any territorial changes without level I consent.

In sum, the plethora of competing level I and level II actors and bargaining dynamics that determine Israeli policy vis-à-vis the West Bank illustrate the explanatory utility of the TLG framework. Level I inputs have consistently exerted pressure to stop Israel from annexing the West Bank (Eiran 2019). Yet at level II, this article assumes that bargaining dynamics will be more mixed, given that the West Bank is perceived as highly valuable (pro-annexation), whilst incorporating its residents is not (anti-annexation). Simultaneously, this article does not assume that inputs at either level will remain constant. Level I actors may exert divergent levels of anti-annexation pressure depending on global political dynamics and the shifting balance of power, for instance. Equally, the growing domestic power of the settlement movement and unilateralist political parties suggests that pro-annexation level II inputs will be stronger in 2020 than in 1967. In both of these cases, this article asks: if the West Bank is so valuable, why were only small parts of it ever annexed?

Partial Annexation: East Jerusalem and the West Bank in 1967

Israel's June 1967 capture of the West Bank ignited an acrimonious debate over the territory's future. The unilateralist Menachem Begin—leader of the right-wing Gahal Party—expressed hope that “Israeli soldiers will yet fly our flag above the Tower of David,” an historical Jewish site in East Jerusalem (Naor 2016, 45). Several similarly unilateralist-minded ministers advocated immediately annexing the entire West Bank. The internationalists leading Israel's broad coalition government, however, objected. Following an extensive debate, Israel's government reached an ambiguous compromise. Israel quietly annexed a small sliver of the West Bank—East Jerusalem and its environs—whilst deferring a decision on the rest of the territory. Why, then, did Israel annex East Jerusalem, but not the rest of the West Bank? Why did Israel annex East Jerusalem in a clandestine manner?

“Liberating” the West Bank: Occupation through a Domestic Lens

Unilateralist and internationalist Israelis alike expressed a yearning for the West Bank, given its plethora of Jewish holy sites—particularly though not exclusively within Jerusalem. Jerusalem's level II intangible value resonated so strongly that Israel ignored level I opposition and declared the city its capital in December 1949.¹ Moving government institutions to the middle of a militarized city—half of which was under Jordanian control—carried a security risk. Nevertheless, Jerusalem's centrality to Jewish Israeli identity overruled these material considerations.

Israel's West Bank occupation exacerbated these feelings of attachment. Prime Minister Eshkol described the war as a “national rebirth” because “Jerusalem was reunited” (State of Israel 1967a). Beyond Jerusalem, Jewish Israelis now had unrestricted access to the West Bank and began to pray at the territory's holy sites, including Hebron's Cave of the Patriarchs and Joseph's Tomb in Nablus. Before June 1967, Jordan prevented Jewish access to these sites. This made Israeli domestic opinion after June 1967 more hard-line (Becker 1971). Propelled by a level II consensus, on June 11, Israel evicted 650 Palestinians who lived next to the Western Wall—one of Judaism's holiest sites—and constructed a plaza for public prayer there (Rempel 1997).

On the same day, the cabinet met to decide whether to annex the West Bank. Israel's wartime unity government saw unilateralists such as Begin share power with their internationalist-leaning Alignment rivals and orthodox Jewish religious parties. Though agreed on little, the unilateralists and internationalists backed annexing East Jerusalem (State of Israel 1967a). Minister Without Portfolio Yisrael Galili argued that “[annexing Jerusalem] is [. . .] worthy even in face of great pressure” (State of Israel 1967a, p. 32). Finance Minister Pinchas Sapir claimed level II values made annexing the city inevitable, since “the nation had already decided, it's in the fibers of its soul” (State of Israel 1967a, p. 34).

Multiple ministers also advocated annexing all or parts of the wider West Bank with significant perceived tangible and/or intangible value. The cabinet agreed that the West Bank's eastern border, the Jordan River, could provide Israel with a natural barrier to prevent an attack from the east, leading Eshkol to declare that “Jordan is the border” (Alon 2017). Labor Minister Yigal Allon proposed that Israel “annex Mount Hebron till the Dead Sea, the Judea Desert, plus greater Jerusalem.” Begin advocated immediate Jewish settlement “in the big cities” with the most salient cultural value (Alon 2017). Significant sections of Israeli public opinion also backed a comprehensive annexation. Ten weeks after the war, over 163,000 Israelis signed a petition against any withdrawal from the territory, whilst July 1967 saw hard-liners found the Movement for Greater Israel, a cross-party organization that employed security and ideational justifications to advocate annexing the entire West Bank (Becker 1971).

On the other hand, Israel's pre-war population stood at 2.65 million citizens, 88.2 percent of whom were Jews. Should Israel annex the West Bank, it would incorporate 900,000 Palestinian Arabs (Della Pergola 2017). Ministry of Foreign Affairs officials argued that annexing the territory would cause, “Grave demographic consequences [. . .] significant economic costs [and] grave cultural and social problems” (State of Israel 1967b). Encapsulating the contrasting domestic inputs, Defence Minister Moshe Dayan argued: “The West Bank [is] Israel,” but “preventing a million Arabs from joining Israel is as important as the West Bank” (State of Israel 1967b). The ministers were also divided over exactly what parts of the West Bank Israel should annex, beyond Jerusalem. Despite this, even relative internationalists, such as Allon, backed annexing large parts of the territory and rejected a return to Israel's pre-war West Bank borders since “In Hebron, Bethlehem, and the surrounding villages, there are about 60,000 Arabs, not such a terrible number [. . .] For an addition of a few tens of thousands, I wouldn't give up Hebron” (Alon 2017).

Thus, level II bargaining explained why Israel annexed East Jerusalem, but does not illustrate why it refrained from implementing the same policy elsewhere. Nor does it explain the clandestine nature of Israel's East Jerusalem annexation. The domestic consensus on East Jerusalem suggested Israel's leaders could maximize their level II gains by loudly proclaiming annexation. Simultaneously, though the cabinet was divided over where exactly Israel should apply sovereignty, it largely agreed that Israel should annex parts of the West Bank beyond East Jerusalem. As such, identifying what stymied this domestic pro-annexation consensus from guiding Israeli policy requires scrutiny of level I bargaining.

Give and take: Mitigating international Pressure

Level I considerations shaped Israel's decision-making in 1967. During its debate over whether to annex the West Bank, the cabinet repeatedly paused its discussions to receive updates from Israel's UN delegation (State of Israel 1967a). Internationalists deemed reports that Jordan, the Vatican, France, and several South American states backed placing Jerusalem under international rule particularly concerning. Jerusalem's intangible value stoked fears the international community would oppose annexation, particularly after US National Security Advisor McGeorge Bundy warned Israeli officials that Muslims and Christians would reject “a Jewish guard” controlling their holiest sites (State of Israel 1967a, p. 67). Ministers recalled how US-Soviet pressure forced Israel to withdraw from the Sinai in 1956 and feared a united front could re-materialize. Israel's UN delegation predicted the pressure would peak at a General Assembly meeting starting June 17 and asked the cabinet to postpone any annexation until afterwards, to prevent agitating international opinion (Friesel 2016).

Yet, the impending UN meeting precipitated the cabinet to agree that annexation should take place sooner, rather than later. Displaying an internationalist disposition, Eshkol urged the government to agree on a collective position since “We will not be able to ignore [the US] for long [. . .] we need to develop answers to the questions the Americans may ask us” (Rosental and Tzoref 2002, p. 679). Most ministers agreed that acting rapidly would deflate level I pressure because any criticism would be post ipso facto (State of Israel 1967a). Thus, Israel's cabinet backed annexing East Jerusalem on June 11, six days before the UN debate.

Simultaneously, Israel offered concessions to avert excessive level I costs. Firstly, the cabinet agreed that the annexation law should not even mention Jerusalem. Instead, the law granted the Internal Affairs Minister the authority to expand any Israeli city's boundaries (State of Israel 1967a). This allowed Israeli UN delegates to claim that East Jerusalem was not annexed and frame the law as purely administrative. This deliberate obfuscation precipitated the long-term debate as to whether, legally, Israel had annexed East Jerusalem. Level I inputs precipitated this policy after US officials advised Israel to “avoid any announcements on annexation” (Alon 2017). Furthermore, Israel's cabinet passed a law guaranteeing freedom of religion, after Bundy's prompting that “You should declare day and night your commitment to the protecting the holy sites” (State of Israel 1967a, p. 67). Israel then instructed its diplomats to downplay annexation and highlight the law guaranteeing protection of the holy sites. Finally, the cabinet agreed that, to avoid a level I backlash, Israel would not remove more Arab residents from Jerusalem's Old City (Alon 2017).

Most critically, Israel combined its unilateralist-leaning annexation policy with an internationalist-appeasing offer to exchange all of the Golan Heights and Sinai Peninsula—territories it captured in the Six Day War—for peace treaties with their recognized sovereigns: Syria and Egypt. Israeli diplomats spun this as an unprecedented offer. However, recent research suggests this was a ruse to mitigate level I pressure (Raz 2012; Raz 2021). The absence of either of these territories in the offer's text belies its purpose. The offer dovetailed with UN Security Resolution 242 of November 22, 1967, which demanded Israeli withdrawal from “territories” occupied in 1967, in return for peace with its Arab rivals. The Arab states then rejected Israel's offer at the Khartoum Summit of late 1967. This allowed Israel to argue it had no peace partner and could indefinitely occupy the West Bank until one emerged (Bregman 2015).

Israel's level I bargaining was not just symbolic: The global balance of power affected its West Bank policies. When fighting broke out in 1967, Israel never envisioned conquering the West Bank, because Jordan was a US-backed power (Raz 2021). Israeli officials felt pragmatism in international affairs trumped the West Bank's tangible and intangible value. Israel's occupation of the territory after Jordan entered the war did not alter this perception. Ministry of Foreign Affairs officials cautioned that any extensive annexation would incur the US’ censure. The US warned that Israel should “act modestly and talk less about territorial ambitions,” leading Allon to admit that “it would be very bad if the world gets the impression that we really intend to hold onto the entire territory” (State of Israel 1967b).

In sum, Israel answered level I anti-annexation inputs through a bargaining game. International dynamics determined the annexation's timing, since Israel annexed East Jerusalem before the General Assembly meeting. Israel also defused the movement to internationalize Jerusalem by promising to protect the city's holy sites, whilst downplaying its de facto annexation. Level I bargaining explains why Israel annexed East Jerusalem in a clandestine fashion, whilst fears of upsetting Jordan and the US disincentivized a wholesale West Bank annexation. Concurrently, Israel's offer to return the Sinai and Golan demonstrates that external inputs affected its territorial policies beyond the West Bank.

Limited Annexation As a Win-Set

Thus, level I and level II bargaining alike affected Israel's territorial policies. Level II pressure made Israel seek to annex large portions of the West Bank. Level I dynamics determined the annexation's timing. It was bargaining between these levels of analysis, though, that determined that Israel would annex East Jerusalem, whilst “deciding not to decide” about the wider West Bank (Raz 2012, p. 44). This topography of annexation constituted a win-set, since it allowed Israel's leaders to maximize level I and level II benefits and minimize any costs.

Firstly, the cabinet carefully coordinated domestic and international inputs when implementing annexation. Most ministers did not participate in the Knesset discussion on annexation. Unusually, the Justice Minister asked Israel's newspaper editors to underreport the legislation. The cabinet then declared its meetings secret, so no transcripts would be released. All these actions sought to avert a level I backlash (State of Israel 1967c; State of Israel 1967d). Israeli leaders balanced pro-annexation domestic pressure and anti-annexation international inputs by applying sovereignty over East Jerusalem, but doing so quietly. This allowed Israel to fulfill level II expectations, whilst mitigating level I opprobrium.

Secondly, it was a two-level game that allowed Israel to annex East Jerusalem. Allon noted that “If we accepted the Arab demands on Jerusalem, we could sign a peace treaty with Jordan tomorrow” (Becker 1971, p. 48). Yet, the ministers agreed that failing to annex Jerusalem would create so much level II dissent it would topple the government (State of Israel 1967a). Decision-makers in the US shared these fears and felt that “the Israelis are emotional to the point of being irrational” over Jerusalem (Raz 2012, p. 70). American officials worried that demanding Israel refrain from annexing the city would generate a level II backlash that would only empower the unilateralists. Since Israeli officials prioritized the US when weighing up potential level I costs, the lack of censure was a critical input in precipitating East Jerusalem's annexation (State of Israel 1967a; Raz 2021).

Thirdly, level I and level II inputs determined what parts of the West Bank Israel included when annexing East Jerusalem. Israel expanded the city's size from 38,000 dunams to 108,000, which former Israeli Prime Minister Ehud Olmert (2019) notes “included all kinds of villages that were never part of Jerusalem and which have no emotional meaning to us.” This new “Greater Jerusalem” appeased unilateralists who wanted to annex the entire West Bank and internationalists who sought to limit the law to Jerusalem. The cabinet also perceived that the US would not tolerate Israel annexing territory beyond Jerusalem. So, Israel enlarged the city to please as many stakeholders as possible (State of Israel 1967a; Alon 2017).

Finally, level I and level I inputs stopped Israel from annexing more of the West Bank. Begin cautioned that “We must not be tempted to chase moderation” and argued Israel could annex the West Bank without granting the Palestinians citizenship (Alon 2017). But other, more internationalist-inclined ministers overwhelmingly disagreed. Police Minister, Eliahu Sasson, claimed: “You can't deprive [the Palestinians] of their rights, not in the twentieth century” (State of Israel 1967e). The Prime Minister's Special Representative for the West Bank warned international pressure would force Israel to grant the Palestinians citizenship, which would “place our future in doubt, as a Jewish state” (Sasson 1967). Simultaneously, Foreign Minister Abba Eban cautioned that the US would censure a comprehensive annexation and described their stance as “Yes to [annexing] Jerusalem, but no to [the West Bank]” (Friesel 2016, p. 387). Encapsulating the link between level II demographic fears and level I opprobrium, Eshkol claimed, “[the Palestinians] will ask why they don't have voting rights, and then we'll have terrible international problems” (Alon 2017).

This perception shaped Israel's emerging policy. Eshkol backed down from annexing more territory and declared that “I don't want more land, and I don't want more Arabs” (Alon 2017). The level II “demographic threat” coalesced with level I norms and power dynamics to bolster the relative internationalists, who then implemented a temporary policy that would prove surprisingly durable, that Israel refrain from annexing more of the West Bank, until it could reach an agreement to divide the territory with Jordan (State of Israel 1967a).

Israel formalized this compromise by loosely adopting the Allon Plan.² The Minister of Labor, Yigal Allon, proposed that Israel annex parts of the West Bank with the highest tangible value, whilst returning the rest to either Jordan or the Palestinians, in exchange for peace. The plan reflected a level II consensus that Israel should neither return to its pre-war West Bank borders, nor incorporate millions of Palestinians. Yet, it did so without immediately annexing more of the West Bank beyond “Greater Jerusalem,” ensuring the plan won the US’ backing. Reflecting the merging of these international and domestic inputs, Allon referred to the plan as “a little peace and a little annexation” (Becker 1971, p. 55). Israel then failed to reach an agreement with either Jordan, or the Palestinians, ensuring the Allon Plan did not constitute a win-set, since any annexation would generate significant level I costs. Instead, Israel opted for Dayan's proposition to “stick to a military rule for a long period of time” (Alon 2017).

Proclaimed, Comprehensive Non-Annexation: The West Bank in 2020

Events in 2020 appeared to herald the end of Israel's deliberate territorial indecision. Writing in May 2020, one commentator noted “annexation [of the West Bank] now seems as likely as at any time since [1967]” (Jackson 2020). But, by early August 2020, annexation “had all but disappeared from the public agenda” (Ahren 2020). This is surprising, given that pro-annexation inputs at both level I and level II were ostensibly more salient in 2020 than in 1967. Prime Minister Benjamin Netanyahu had openly declared he would annex all or most of the West Bank. The US had tempered its opposition to annexation and, in 2019, had recognized Israel's annexation of another contested territory, the Golan Heights. Subsequently, Netanyahu's Likud Party won the largest vote share in national elections on a pro-West Bank annexation platform and formed a heavily unilateralist-leaning government (Levinson 2018). Why, then, did Israel's leaders fail to fulfill their declared annexation attempt?

The Rise and Fall of an idea: Domestic politics and the Annexation Plan

Conforming to this article's assumptions, the origins of Israel's 2020 annexation attempt lie within level II bargaining. With a unilateralist-leaning Likud-led coalition in power since 2009, pressure had long been building for Israel to annex the West Bank. In 2017, right-wing NGOs precipitated a surge of pro-annexation Knesset legislation. In 2018, the Likud Central Committee unanimously backed annexation (Levinson 2018; Lynk 2019). Annexation was popular with Likud voters, whilst polls in 2019 indicated increased public support for the policy (Kraft 2019). Netanyahu had earlier called for a two-state solution and had blocked pro-annexation bills. In April 2019, by contrast, Netanyahu claimed a Palestinian state would “endanger our existence” and instead backed annexation (Haaretz, 2020).

Israel's four inconclusive elections within two years fueled the pro-annexation momentum. Competing for unilateralist support, rival parties sought to outbid each other in their pro-annexation stances. Initially, Netanyahu suggested he would annex the three largest settlement blocs closest to Israel's border. But after the rival New Right party pledged to annex around 60 percent of the West Bank in the April 2019 elections, Netanyahu went even further before the September 2019 election and promised to annex every West Bank settlement, as well as the entire Jordan Valley—the territory's eastern border (Gur 2020a). There were also level II institutional issues influencing policy. By 2019, Netanyahu deeply distrusted Israel's anti-annexation security and foreign policy bureaucracy, because he saw them as part of the establishment that sought to remove him from office through an ongoing corruption trial. Feeling besieged, Netanyahu became more receptive to relatively hard-line unilateralist advisors (Ravid 2022). Simultaneously, Netanyahu's stance reflected public opinion: the Likud returned as the largest right-wing party in the 2019 elections, whilst polls suggested only 30 percent of Israelis rejected annexing all or parts of the West Bank (Reuters 2019).

When Netanyahu finally formed a government after the March 2020 elections, an elite level II consensus backed annexation. Israel's government was more ideologically diverse than the overwhelmingly unilateralist previous Netanyahu-led administration in that it contained the Blue and White Party. However, even the internationalist-leaning Blue and White backed a limited annexation of parts of the West Bank with the most perceived tangible value: The Jordan Valley and the three major settlement blocs (Ravid 2022). Subsequently, all parties within the coalition pledged to suspend legislation unrelated to the COVID-19 crisis, with one exception: Annexation. The coalition agreement allowed the Likud to advance annexation even without Blue and White's support from July 1, 2021. Within the legislature, Netanyahu would not require the relatively internationalist party's backing, since 68 of the 120 Knesset members were from pro-annexation unilateralists and Netanyahu-allied religious parties (Levy 2020).

Nevertheless, after the new government took office in mid-May 2020, the consensus which had propelled Israel's annexation attempt rapidly dissipated. By early June, only 34 percent of Israelis supported annexation; even most unilateralists no longer backed the policy. With the COVID-19 pandemic causing unprecedented disruption, only 5 percent of voters felt annexation constituted a national priority (Gur, 2020a). In late June, Blue and White's leader, Benny Gantz, claimed: “Dealing with the coronavirus and its socioeconomic and health consequences is the more pressing issue” than annexing the West Bank (Bachner and Magid 2020).

In sum, level II factors engendered and then stymied Israel's annexation attempt. Domestic politics first exerted pro-annexation pressure. Israel's elites responded to lobbying by their unilateralist base to legitimize annexation and the policy became increasingly mainstream. Israel's internal political turmoil exacerbated this pressure. Subsequently, a level II backlash undid this tentative pro-annexation consensus. This hemorrhaging of support partly explains why Israel aborted its annexation attempt. Israel's leaders lacked sufficient level II incentives: They would yield few benefits from the policy and had little reason to implement it. It does not explain, though, what factors precipitated this significant shift in level II opinion.

From Apathy to Carrot and stick: Global responses to Israel's annexation Attempt

Mimicking Israel's domestic politics, level I power dynamics initially furthered annexation. The Trump administration had already broken with decades of US policy by recognizing Israel's Golan Heights annexation. Israeli unilateralists argued Trump's tenure represented an unprecedented opportunity to annex the West Bank, given any censure would be minimal (Satloff 2021). This claim was bolstered in January 2020 when the US published a peace plan that suggested Israel could annex every West Bank settlement and around 30 percent of the territory (Ravid 2022). The US’ supportive stance coalesced with global apathy: distracted by the COVID-19 pandemic, responses to Netanyahu's plan were muted.

But, as Netanyahu's declared annexation deadline of July 1, 2020 approached, regional actors coordinated to demonstrate that the policy would irreparably harm Israel's security and foreign relations. Qatar threatened to cut off the millions of dollars it provides to the Gaza Strip, which would invalidate existing agreements between Israel and Hamas—the territory's Islamist rulers—and risk provoking conflict (The Times of Israel 2020). In May, the Palestinian Authority (PA) ended its long-term security cooperation with Israel (Baconi 2021). Jordan's King Abdullah threatened a “massive conflict” if annexation were to occur (Al Jazeera 2020). This burgeoning level I pressure filtered down to level II. Speaking to serving soldiers in June 2021, the Israel Defence Forces Chief of Staff, Aviv Kochavi, warned: “In a few weeks, you may find yourself in [the West Bank] because of riots and terror” (Fulbright 2020).

Level I actors with whom Israel enjoyed robust relations cautioned against annexation. In early July, Josep Borrell, the EU Representative for Foreign and Security Affairs, advised that the policy would harm ties with the EU, Israel's largest trading partner (Baconi 2021). Germany's Foreign Minister traveled to Jerusalem to lobby against annexation; several European states even advocated sanctions against Israel (Ahren and Fulbright 2020). The UK warned that any West Bank annexation would cause it to unilaterally recognize Palestinian statehood. Within the US, the powerful pro-Israel lobby group AIPAC declared it had no objection to its supporters in Congress criticizing annexation. The Democratic Party's entire leadership and most of its representatives in both houses of Congress then denounced Netanyahu's plan, including the party's presidential nominee, Joe Biden (Harris 2020).

This dramatic shift in level I dynamics then re-shaped the Trump administration's stance. In early 2020, Secretary of State Mike Pompeo declared any West Bank annexation was an internal Israeli matter. In mid-May, though, Pompeo demanded Israel only annex territory with the US’ consent. Both sides then formed a committee to map what parts of the West Bank Israel could annex (Halbfinger and Rasgon 2020). By late June, Israeli government sources blamed the US for indefinitely delaying annexation. US officials told the Israelis to be patient because their involvement would lessen global criticism (Bachner and Magid 2020). Israeli officials later complained that, “The Americans promised to bring on board the Arab world. The fact is, they did not” (Caspit 2020a). Israel expected the US’ backing for an immediate, comprehensive annexation. But, the global anti-annexation backlash—especially furtive criticism and lobbying from pro-US Arab Gulf states—caused the Trump administration to seek to temper the plan's pace and scope (Kushner 2022).

Whilst global pressure coalesced to scupper Israel's annexation attempt, level I bargaining produced a ground-breaking incentive: normalization between Israel and the United Arab Emirates (UAE). Rumors had long abounded of clandestine ties between Israel and the Gulf Arab states. An Israeli government source argued annexation would embolden this process since “Our friends in the Arab world are cheering us on” (Satloff, 2021, p. 9). The UAE undermined this assessment by offering full normalization on the condition that Israel abandon its annexation plan. Israel agreed to what later became known as the “The Abraham Accords” in early August. The proposal came from the UAE. But the US negotiated the contours of the agreement with the UAE, before offering it to the Israelis and acting as an intermediary (Kushner 2022; Ravid 2022). Despite Netanyahu's claim that the agreement only postponed annexation, Israel's government then stopped pushing the policy and instead emphasized the benefits of normalization (Makovsky 2020).

Overall, then, level I inputs played a crucial role in enabling and then stopping Israel's annexation attempt. At first, level I criticism was minimal and emboldened level II unilateralists. In mid-2020, though, this dynamic transformed to suggest that annexation would damage Israel's foreign relations and produce a salient national security threat. The level I backlash led the Trump administration to delay annexation, ensuring Israel could no longer implement it unilaterally without incurring significant costs. When the US failed to mitigate international criticism, it offered a choice of either normalization with the UAE, or annexation. This proposed inducement would reverse the international dynamic: Israel would receive tangible benefits for implementing normalization, rather than the substantial costs after fulfilling its annexation attempt. It was this combination of carrot and stick that disincentivized annexation and led to the policy's indefinite postponement.

Non-annexation As a Win-Set

Unpacking level I and level II bargaining partly illustrates why Israel did not annex the West Bank, but it raises several related questions: What spurred the sudden escalation in level I criticism? How did the involvement of the US alter the format and feasibility of the annexation attempt? Why did pro-annexation level II inputs falter? How did level I and level II dynamics interact to engender normalization, instead of annexation?

Firstly, it was level II political dynamics that precipitated the fluctuations in level I opinion and lobbying vis-a-vis annexation. Perceiving the policy was a unilateralist bluff by Netanyahu to win support in the country's four elections—rather than a genuine annexation attempt—the EU, Egypt, and other key actors muted their criticism (Emmott et al 2020; Winter 2020). Israel's electoral turmoil also delayed annexation, since the Knesset—which would have to approve the policy—was rarely in session. The instability ended once Netanyahu formed a government with annexation enshrined in the coalition agreement. This provoked the surge in level I criticism. Furthermore, though the coalition agreement permitted annexation, it was conditioned on three factors: The US’ consent; that Israel consult its international partners; and that annexation not harm Israel's regional relations (Gur 2020a). These conditions, imposed at level II by the internationalist Blue and White, empowered level I actors with the agency to foil Netanyahu's annexation attempt.

Secondly, the US’ participation from mid-May 2020 re-shaped Israel's annexation attempt. Though the coalition agreement stipulated Netanyahu did not require Blue and White's support, it did demand the US’ backing. The US then insisted any annexation achieve a cabinet consensus in Israel (Mualem 2020). As such, level I and level II dynamics empowered the US and Blue and White, both of whom demanded Israel limit annexation to the three largest settlement blocs, whilst yielding 6.5 percent of the West Bank to Palestinian control. Accordingly, Netanyahu changed the plan to restrict annexation to the three blocs and cede 0.5 percent of the West Bank to the PA. This concession failed to blunt level I criticism. The US rejected Netanyahu's 0.5 percent offer as insufficient, whilst Jordan and the PA would not accept any annexation whatsoever, regardless of Israel's concessions (Al-Omari 2020; Caspit 2020b).

Thirdly, it was this combination of level I and elite level II pressure that altered Israeli public opinion. Polling in June showed only 25 percent of Israelis backed annexation without the US’ support, and just 26 percent backed the policy if it endangered relations with Jordan (Hermann and Anabi 2020; The Jerusalem Post 2020). Conversely, the plan's truncation incensed influential Israeli unilateralists with links to the government. By mid-June, half of the West Bank's settler leaders opposed the plan, due to its limitation to the major settlement blocs (Caspit 2020c). June 2020 saw the UAE's US Ambassador, Yousef al-Otaiba, publish an anti-annexation article in the Israeli press. The op-ed, the first time a serving Gulf diplomat had published it in an Israeli paper, asked the public to choose “either annexation or normalization” (al-Otabia 2020). Whilst Blue and White and the US collaborated to make the scope of annexation smaller, inducements and coercive diplomacy led to Israel's public questioning the policy's utility, whilst undoing the plan's unilateralist support.

In sum, it was the complex interaction of level I and level II dynamics that ensured that Israel's pursued normalization over annexation. Netanyahu faced unreconcilable level I and level II demands, with the international community, the US, Israeli public opinion, and his coalition partners demanding minimal annexation, or none at all. The Likud's base, its unilateralist rivals and the settlement movement, however, backed a comprehensive annexation. This ensured that any annexation would yield significant criticism and alienate key level I or level II stakeholders. By contrast, 77 percent of Israelis backed normalization over annexation (Gur 2020b). The Abraham Accords precipitated further normalization agreements with Bahrain, South Sudan, and Morocco. An anonymous Israeli senior official argued that “[Netanyahu] was forced to choose peace after realizing he could not achieve annexation” (Caspit 2020b). In fact, it was a combination of anti-annexationist level I and level II inputs which ensured that Netanyahu fumbled and prevaricated, until he was able to replace a policy with significant potential international costs and declining domestic returns (annexation), with one enjoying broad support both inside and outside of Israel (normalization).

Testing Assumptions: 1967 and 2020 Compared

Despite the divergent outcomes in 1967 and 2020, table 2 below illustrates a striking commonality: the distribution of pro- and anti-annexationist inputs at level I and level II was functionally identical in both cases. Conforming to the TLG literature, level I primarily exerted anti-annexation pressure. As the non-great power literature suggests, the level I balance of power (I_c)—particularly pressure from a superpower (the US)—coalesced with anti-annexationist hegemonic values in the international system (I_b), to constrain a non-great power (Israel). These factors coalesced to suggest that at level I, the impact of annexation would be costly and involve substantial international opprobrium (I_d). The only consistently pro-annexationist level I input was the territory's tangible value (I_a): In both 1967 and 2020, ministers—including relative internationalists—argued for “strategic depth.” This policy necessitated the annexation of at least the Jordan Valley and the West Bank's western border, to prevent cross-border infiltrations. As such, annexation would deprive Israel's state-level Arab adversaries of a topographical, operational or strategic advantage in any future conflict.

Table 2.

level I and level II bargaining in 1967 and 2020

	1967	2020
Level I Dynamics
Territory's tangible value (Ia)	Pro-annexation	Pro-annexation
Hegemonic values (Ib)	Anti-annexation	Anti-annexation
Balance of power (Ic)	Anti-annexation	Anti-annexation
Impact of annexation (Id)	Anti-annexation	Anti-annexation
Level II dynamics
Territory's intangible value (IIa)	Pro-annexation	Pro-annexation
Hegemonic values (IIb)	Mixed	Mixed
Balance of power (IIc)	Pro-annexation	Pro-annexation
Impact of annexation (IId)	Anti-annexation	Anti-annexation

By contrast, level II demonstrated a more nuanced distribution of pro- and anti-annexation inputs in both 1967 and 2020. The territory's intangible value (II_a) was a powerful pro-annexation input, given the West Bank's designation as “Judea and Samaria,” and framings of Jerusalem as “the external capital of the Jewish people” (Zellman 2020, p.343). Equally, the level II balance of power (II_c) was pro-annexation: Israel's governments in 1967 and 2020 contained unilateralists and internationalists, though these actor-types did not disagree on whether to annex or not. Instead, they were divided by the extent of Israel's annexations and whether level I actors should be co-opted or ignored. Conversely, a lone but powerful input—maintaining a Jewish majority within Israel—ensured that hegemonic values (II_b) exerted mixed level II pressure: unilateralists and internationalists valued the West Bank, but many felt that the potential changes to Israel's governing principles and values following any annexation constituted an existential threat. This suggested that the perceived net impact of annexation (II_d) would necessitate a fundamental reconsideration of Israel's governing principles. Table 3 below presents the same inputs as above and demonstrates how, in both cases, anti-annexationist level II inputs coalesced with level I inputs to avert a comprehensive annexation.

Table 3.

bargaining across both levels in 1967 and 2020

Year	Pro-annexation	Anti-Annexation	Mixed	Result
1967	Ia, IIa, IIc	Ib, Ic, Id, IId	IIb	Partial Annexation
2020	Ia, IIa, IIc	Ib, Ic, Id, IId	IIb	No Annexation

The above findings illustrate that analyzing level I or level II bargaining in vacuo does not explain the divergent outcomes in 1967 and 2020. Instead, it was interactions between both levels of analysis—as the TLG framework's assumptions posit—that determined the variance in outcomes, or whether Israel annexed conquered territory or refrained from doing so. In 1967, contrasting level I and level II pressures determined the topography of annexation and the clandestine tactics Israel used when applying sovereignty over a new “Greater Jerusalem.” Israel's leaders were wary of provoking a level I backlash through an extensive annexation that would generate intolerable pressure on Israel to return all or most of the West Bank (I_d). But they also felt that no annexation at all would imperil the government at level II. Unilateralist level II inputs (II_a, II_b, and partially II_c) manifested at level I, convincing the US that lobbying against any annexation whatsoever was untenable. As a result, Israel's win-set became a limited annexation, thereby mitigating level I criticism (I_b, I_c and I_d) and level II demographic concerns (II_b), whilst appeasing hard-line public opinion (II_a and II_b).

In 2020, by contrast, global criticism was initially muted, then increased, ensuring that I_c and I_d transitioned from relatively pro-annexation to staunchly anti-annexationist inputs. Whilst the domestic balance of power (II_c) remained pro-annexation, the US/Blue and White partnership amplified anti-annexationist inputs. The US’ involvement and its interactions with Blue and White re-shaped and truncated the plan. This, in turn, created dissent in Netanyahu's unilateralist base. The result was that either a comprehensive annexation or a limited one would generate an international backlash or imperil Israel's governing coalition (I_d and II_d). Ironically, it was the increased salience of pro-annexation level II pressure (II_a, II_b, and II_c), when compared to 1967, that explains why Israel did not annex any territory in 2020.

The UAE's offer allowed Israel's leaders to replace annexation with another game-changing initiative, the “Abraham Accords.” Though a level I input, the UAE's offer targeted Israel's domestic arena. It sought to alter Israelis’ cost/benefit analysis through incentivizing normalization and correspondingly dis-incentivizing annexation, whilst avoiding the demographic and governance issues and international condemnation that annexation would bring (due to I_b, I_c, I_d, II_b, and II_d). This initiative succeeded and caused Israel's leaders to perceive normalization as an optimal win-set over annexation. In sum: analysing one level of analysis alone does not highlight the causal paths that explain the two divergent outcomes across both cases. Instead, it was the interaction of level I and level II dynamics that determined Israeli policy. This illustrates the utility of the TLG framework for explaining the dissimilar outcomes in 1967 and 2020: partial annexation and non-annexation.

Conclusions

Both the cases analysed here at first appear distinct. In 1967, Israel partially and clandestinely annexed the West Bank, following its victory in a major regional war. In 2020, Israel sought to end a decades-old West Bank occupation through a declaratory, comprehensive annexation. These divergences mask a fundamental commonality: in both 1967 and 2020, Israel's leaders secured their win-set by navigating a complex bargaining equation spanning level I and level II inputs. Whilst international (level I) inputs were overwhelmingly anti-annexation, domestic (level II) inputs were more divided. In both cases, coordination between anti-annexationist level II and level I actors prevented Israel from annexing all or most of the West Bank.

This article vindicated the claim made by TLG scholars that coalition governments increase level I inputs in level II bargaining since leaders prioritize the demands of coalition partners (Alons 2007). Unilateralist coalition partners act as “veto players,” employing their level II inputs to scupper level I negotiations (Mo 1995, p. 914). Level II intransigence can bolster a state's level I negotiating power since “where popular support exists for sustaining contentious claims, political elites are empowered internationally to portray them as non-negotiable” (Mo 1995; Zellman 2015, p. 492). Accordingly, states pursue “reverberations,” intervening in a level I bargaining partner's level II dynamics to re-shape its domestic politics.

This article illustrated that each of these assumptions manifested, but often in more complex manners than the literature suggests. Coalition dynamics—unilateralist-internationalist tension in particular—did constitute a critical level II input. But, rather than empowering level II inputs at the expense of level I bargaining, Israel's multi-party governments facilitated coordination between both levels of analysis. Much of the TLG literature simplifies level II inputs as “a constraining factor” that scuppers a level I agreement (Hurst, 2016, p455). But, whilst demonstrating the TLG framework's utility for understanding Israeli policy, this article found no evidence for reductionist claims that international and domestic politics exclusively disincentivize or incentivize annexation, respectively. Rather than impose this assumption, future studies of non-great power annexation attempts should instead acknowledge the complex inputs at play in both levels and chart the interaction between them.

In 1967, the US recognized hard-line level II attitudes and steered Israel towards limited annexation rather than pressure for no annexation whatsoever. This suggests that unilateralist level II inputs empowered Israel to exert level I intransigence. In 2020, the inverse manifested: though elite and public opinion backed annexation, Blue and White's conditions empowered level I actors. Equally, the COVID-19 pandemic dramatically shifted Israel's public and elite policy priorities away from annexation. Thus, “reverberations” constituted multiple, divergent actors in two different levels of analysis empowering each other, rather than a consistently straightforward intervention in a state's level II bargaining by a level I actor. Further, reverberations are not always actor-led: “black swan” events such as the COVID-19 pandemic can re-shape systemic and unit-level balances of power (Tinh and Negan 2022, p. 303).

Most critically, the 1967 case exemplifies that annexation attempts are not like international summits; the former have more potential outcomes than success and failure (or annexation and non-annexation). Since Putnam's original work, subsequent studies have applied the TLG framework to explain a range of phenomena beyond summit diplomacy. This article vindicates this approach, but it also shows that the future studies should not impose a success/fail outcome format where this binary represents an oversimplification. Overall, this article joins the minority of TLG studies that reject mono-causal chains and instead illustrate the reflexivity of inter-level bargaining and multiple potential outcomes).

Alongside its implications for the TLG literature, this article contributes to understanding annexation attempts. Further research on annexation is exigent. In 2008, one scholar predicted that “to the degree that conquest and annexation are undesirable behaviors, international relations may soon take a turn for the worse;” a claim that Israel's West Bank annexation attempt vindicated (Fazal 2008, p. 238). On the one hand, Israel mimicked the great powers scrutinized in the literature since it pursued annexation due to the West Bank's tangible and intangible value. On the other hand, Israel—a non-great power—behaved and was influenced differently than great powers. Whereas level II bargaining exerts the most agency in great powers, by focusing on Israel, this article affirmed that bargaining between levels I and II is crucial for determining the outcome of non-great power annexation attempts.

This article partly affirms the literature on annexation. On the one hand, it confirmed that decision-makers consider a policy's utility for advancing their domestic interests. They also illustrate that demographics are critical in determining the topography of annexation, because demographic changes cause corresponding shifts in a state's identity and governing principles (Maass 2020). On the other hand, these findings contradict claims that domestic politics matter more than international politics during annexation attempts. It also illustrates that non-great powers behave differently than the great powers analysed in the annexation literature. Israel was repeatedly constrained by level I pressure, even when the conditions exist that empower level II inputs, such as coalition governments, unilateralist domestic opinion and the exceptionally high perceived tangible and intangible value of the territory itself.

Nevertheless, this article scrutinized one milieu alone. Further, the limited agency of the territory's residents—the Palestinians—in affecting Israel's external and internal bargaining manifested across both cases but did not correspond to the broader literature on territorial conflict, which stresses that the occupied population frequently affects decision-making within the occupier (Edelstein 2008). Future works should apply the TLG framework to other annexation attempts where bargaining with the occupied population constitutes a level I input. Additionally, the TLG and territorial conflict literature suggests that significant variation exists between democracies and non-democracies (Thu and Tinh, 2023). Thus, future research could employ this article's findings to comparatively assess the TLG's utility for understanding non-democratic, non-great power annexation attempts and democratic non-great power annexation attempts beyond Israel.

This article delineated findings relevant to the contemporary security problem of how to prevent occupiers from fulfilling annexation attempts. Writing in 1937, George Orwell argued that “No modern man, in his heart of hearts, believes that it is right to invade a foreign country and hold the population down by force” (Orwell 1937, p. 126). Nonetheless, annexations continue to challenge the contemporary world order and the norms underpinning it. Israel's policies in 1967 demonstrate that global condemnation alone is unlikely to stop a limited annexation that is conducted stealthily, rapidly, and with overwhelming domestic support. Conversely, this case showed that international opprobrium and global norms upholding self-determination can play a role in geographically stymieing annexation attempts, despite domestic support for the policy. The 2020 case reveals that significant incentives can deter annexation, since the “Abraham Accords” were the most critical international game changer. In both 1967 and 2020, external pressure and lobbying empowered annexation skeptics within the occupier and elicited a policy change. In the West Bank or elsewhere, annexation can be averted or tempered, but this requires careful coordination between the domestic and international levels of analysis and a robust level I response.

Funding

Rob Geist Pinfold acknowledges funding from Charles University, grant UNCE/HUM/018 - Peace Research Center Prague.

Footnotes

1

See UN General Assembly Resolutions 181, 194, and 303.

2

Though the plan was never adopted as formal Israeli policy, it was the closest the cabinet came to an informal consensus (Bregman 2015).