Existing approaches: why it is not oil, imperialism, or repression
Since monarchies are rare and concentrated in one region, they share a plethora of other characteristics apart from their political system, and to disentangle their effect from monarchism is a challenge. How do we know that it is not merely one of these similarities that explains monarchic foreign policy behavior instead of the political system?
Complementary’ confounding factors: geography and kinship
Especially the monarchies of the Persian Gulf share a broad range of commonalities - geographical proximity and institutions; a common culture and historical trajectory; and, in most cases, an economic system. However, while they pose problems for the analysis, because they complicate the isolation of the effect of the political system per se, they are complementary rather than competing explanations. This is because they are not necessarily a separate factor, nor do they always have the same direction of effect as a SIT-based mechanism.
Except for Oman,19 all of the Gulf monarchies are dynastic (see Herb 1999), meaning family dynasties form the ruling class and the pool of higher-level political personnel. These dynasties are further connected with each other via various marriage, family, and tribal relations and shared common identities while forming a regional subsystem constituted by kinship and alliance.
Far from being an alternative to a SIT-centered explanation, these shared features make monarchies are a likely case of it. Family and tribal ties are generally more likely to be found among monarchies given that intermarriage is rarer between or with republics. The especially close bond that the Gulf states share intensifies the ingroup identification among the members of the subgroup of the Gulf monarchies as forms part of their shared similarity. The commonalities of the Gulf monarchies would thus not constitute a separate explanatory factor but rather form part of the monarchic traits of the system. Also, close ties alone have not discouraged countries from waging war against one another in the past, as the prewar (pre-world-war) European monarchies plainly show.
This also applies to geographical clustering, which makes the monarchic peace even more puzzling, because it should raise the possibility of war. While geographic proximity or contiguity can have an ambivalent causal effect on war and peace, for most except the major powers, it is a necessary condition for a war to occur (Diehl 1985) and at least contributes to war (Diehl 1991) - which is why it is recommended as a control variable to avoid overestimating the probability of war. Nonetheless, although it may be necessary, it is not sufficient for war. Military escalation is not inevitable for neighbors, if they find alternative ways to solve their conflicts (cf. Vasquez 2009, 161). Therefore, concentrating the study on the Middle East is the most promising course of action - not only has it the largest concentration of remaining authoritarian monarchies, providing them with opportunities for conflict absent between more distanced monarchies, but it also provides an ample concentration of another rare phenomenon: interstate war. Between 1948 and 2003, 11 such wars occurred here.
Rival confounding factors: small states, oil, alliance and coup-proofing
There are other traits that cluster among Middle East monarchies that might make them more peaceful without being linked to their political system per se. The first such commonality is their size. Most monarchies are also small states, especially if measured by citizen population.20 Small states have only limited foreign policy options at their disposal compared to larger and thus generally more-powerful states (Baker Fox 1959; Keohane 1969; Peterson 2006). Two main expectations for foreign policy derive from the limitations that small states (or micro-states) face (cf. Peterson 2006):
Both might preclude (offensive) war. A situation wherein Qatar attacks Saudi Arabia, more than 14 times its size in population, is almost inconceivable, even during the tensions since the beginning of the blockade. One the other hand, the reverse is possible in such a case, making a small and wealthy statelet like Qatar a low-hanging fruit to any larger states to pick, including other monarchies, thus raising the possibility of military conflict. However, most of the broken-down monarchies, as well as Jordan and Morocco, are medium or even large territorial states. In addition, many studies have shown that small states can still conduct a varied and confident foreign policy, including conflict participation (see e.g. Al-Khalili 2009 for the case of Oman; and Kamrava 2013 for the case of "smart power” Qatar).
Apart from belonging to the same subregion and mostly exhibiting small-state characteristics, the Gulf monarchies are also rentier states (Beblawi and Luciani 1987; M. Beck 2009) with a high dependence on resource export and the ensuing high income. Analogous to a common critique of the DPT (Gartzke 2007), this fact suggests that the driving force of external behavior is not regime type but rather economy or wealth: rich states tend to be samrated status quo powers, i.e. “states that seek only to keep their resources and for whom the costs of war exceed the gains from nonsecurity expansion” (Schweller 1997.46), and pursue a comparatively more conservative and thus risk-averse policy. In addition, external powers dependent on the exported resources have a high interest in the protection of the exporting states.
However, oil is not a necessary explanation for peacefulness, as the existence of the resource-poor monarchies Jordan and Morocco as well as their former brethren Egypt and North Yemen in their monarchical periods show. It is also not sufficient, as Iraq under Saddam Hussein repeatedly demonstrated. Some of the most "radical” or confrontational states in the high times ofpan-Arabism and antiimperialism were rentier states like Algeria, Libya, Iraq, and Iran.21
On the contrary, oil can lead to conflicts because the wealth-induced heightened military capabilities raise the probability of military success and therefore also the probability of war (Gause 2010). Furthermore, the search for oil can lead to more territorial conflicts, of which most wars encompass at least an element of (Vasquez 2009, 166). Thus, there is no clear causal expectation for rentierism or oil wealth. Fred Halliday even asserts that “oil has no significant causative relation” to most interstate conflicts in the region (2011, 271, FN 31).
In addition, the state of saturation can be achieved only if the allocation and cooptation of societal groups enabled by the resource export (Ross 2001) works. Changing oil prices can affect this ability. If oil wealth were the defining factor for monarchical peacefulness, boom and bust periods would directly relate to foreign policy behavior. Instead, we find that monarchic war involvement is consistently low and peaks only in the First Gulf War, when (oil-rich) Iraq attacked monarchic (and even more oil-rich) Kuwait and the fellow monarchies flocked to its aid. In addition, if oil wealth ameliorated conflict, the expectation for the resource-poor monarchies of Jordan and Morocco would be higher war-proneness. Although those are also the only monarchies to initiate conflict, they also never fought against other monarchies - in contrast to oil exporters Iran and Iraq, which did fight each other.
Due to their lesser wealth, Morocco and particularly Jordan are highly dependent on the external support by Western powers, but this is a commonality that all monarchies share to some extent. This phenomenon even holds true on a global scale, as every monarchy since the Second World War has been a Western ally, but it is even more salient for the Middle Eastern monarchies that were mostly created, or at least shaped, by the British.
Alliance structure is an important determinant of foreign policy behavior, and a possible monarchical peacefulness could therefore be an artifact of the Cold War and the bloc structure. If all monarchies are Western allies, there should be no need or opportunity to attack one other. Furthermore, they might feel safer because of the promise of external protection and therefore less prone to preventive or preemptive wars. (Defensive) alliance ties are also able to deter potential aggressors (Leeds 2003). This is a strong argument, especially because it is dyadic in nature with the outcome of depending on the alliance affiliation of both members of a dyad.
Nevertheless, it is not sufficient grounds for dismissing other theoretical approaches. The extent of Western support differed over time and was never an absolute determination for the choice of foreign policies by Middle Eastern rulers. Even close Western allies show significant autonomy. This is most drastically emphasized by Jordanian refusal to fight Iraq despite pressure from the whole Western bloc and most of its own Arab neighbors in the First Gulf War of 1990-1991, which also illustrates the failure of deterrence by alliance. In addition, Saddam Hussein’s invasion of Kuwait as well as Arab-Israeli wars show other instances when Western alliance alone was not a sufficient explanation to keep two (or more) Western-backed states from waging war against one another. Furthermore, ally entanglement in wars can lead to higher conflict participation (cf. Christensen and Snyder 1990).
There is sufficient ground to argue that Western alliance is in fact not so much an explanation for the foreign policy behavior of monarchies as it is the outcome of the same properties of monarchies that also account for other features of the foreign policy behavior of monarchies. Monarchies are almost always Western allies, an anomaly that can be explained in part by monarchic specificities making monarchies particularly suitable for alliance with external Western powers
(cf. Sunik 2015). It thus cannot be considered separately from factors relating to the political system of these states.
When thinking about what factors could account for a difference in conflict initiation in the Middle East, the role of the military stands out. Due to the historic threat of military coups - in fact, four of the five post-WWII MENA monarchies have been removed by military coups d'état - the surviving monarchies tried to coup-proof their regimes by fragmentation, redundancy, and narrowing the loyalty bounds of their militaries (Quinlivan 1999). Power-sharing and balancing purposes were additional drivers of these processes that weaken military power in the monarchies. In Saudi Arabia under King Abdallah, for instance, the king ran the National Guard, Crown Prince Sultan controlled the air force and was minister of defense, and Crown-Prince-in-waiting Nayif commanded the interior forces (Billingsley 2009, 64).
A possible conclusion would be that since the military is not an integral part of the regime, military options become less likely. Since leaders with a military background tend to be more war-prone (Horowitz and Stam 2014), this could explain the difference between the republican presidents who mostly share a military background and monarchs who usually (but not universally, as King Hussein of Jordan or Sultan Qaboos of Oman demonstrate) do not. Military socialization normalizes war as an option and tilts the cost-benefit analysis of employing military violence abroad while pacifist leanings are often penalized in a military junta context (Brecher 1996; Debs and Goemans 2010; Huntington 1957; Posen 1986). Therefore, if the military is part of the regime, military options might be more widespread. Civilian monarchies would thus not be more pacific per se. Instead military republics would be more belligerent, which would account for the monarchy-republic divide.
There are a few caveats to this explanation, however. First of all, there are studies disputing a greater bellicosity in military personnel (Betts 1991; Feaver and Gelpi 2005; Sechser 2004). Not only might military leaders not differ much from civilian leaders in war-proneness, but also Martin shows in his extensive analysis of the South American "zone of peace” between military republics that common military socialization can in some circumstances even impede armed conflict (Martin 2006).
Even if we concede a greater propensity for military conflict in military regimes, other objections remain: coup-proofing was not implemented to the same extent in all monarchies - in the linchpin monarchies of Jordan and Morocco, the military played a greater role than in the dynastic monarchies. There were purges following military unrest and coup attempts there, but the military was never neutralized like in the Gulf states and continues to form an important pillar of the stability of the regime, especially in Jordan (cf. Pollack 2002). Also, coup-proofing was never an exclusive monarchic tactic; on the contrary, it was most pronounced in republican military regimes such as Iraq and Syria, where coup-experienced leaders saw enough iterations to gain in-depth knowledge on how to prevent them and used that knowledge to stabilize their own rule (Heller 1977; Lutterbeck 2013). The
Moving beyond democratic exceptionalism 43 belligerence of military regime would thus at most be an incomplete explanation for a monarchic peace.
In summary, it appears that most shared characteristics of monarchies in the Middle East discussed earlier are static and, though they certainly can explain some events, cannot explain dynamics and changes. With the exception of Western alliance, the majority of these explanations is monadic and can also explain only war or peace, but not both. In contrast, a SIT-centered explanation is dyadic and can explain changes over time as well as wars with non-monarchies and peace among monarchies.
A more comprehensive explanation always encompasses theoretical synthesis to an extent. The aforementioned factors have been covered at length in previous works. They are not dismissed for the benefit of a SIT-centered explanation but complement it.
But what can we say about other monarchy-related explanations of their conflict behavior? Although monarchy research has rarely explicitly aimed to explain foreign policy behavior, there are some indications that monarchies are special in numerous ways.
Monarchic pacifism from an institutionalist perspective
Monarchy research is a valuable source of explanation for a monadic effect of monarchism - i.e. if there is something about monarchic systems in general that generates greater peacefulness or lesser belligerence in contrast to other types of systems, such as juntas, personalist dictatorship, or single-party systems, it should be found here. What sets monarchies apart from other political systems are specific monarchic institutions - most obviously the position of the head of state, the monarch. The nature of the institution of the monarch enables a greater pragmatism, especially in foreign policy, and restricts ideological chain ganging and brinkmanship.
Despite the limiting concentration on the survival of monarchical authoritarianism, research on the Middle Eastern monarchies bears many fitting leads for research on their foreign policy as well. Lisa Anderson (see e.g. 2000) points to the special advantages of monarchies which are not bound to a specific overreaching ideology, an aspect that she attributes to the prerequisites of state building. Whereas presidents are per definition always part of a particular political group or party and - especially in the Arab world (in religious terms also in Iran) - came to power through ideology-driven revolutions, monarchs, whether they be kings, emirs, or sultans, are formally above everyday politics and mostly do not have a pronounced political ideology, in particular not one directed externally (Anderson 2000; Lucas 2004; Waterbury 1970).
Monarchy also brings advantages for legitimation. Many monarchies allow for parliaments occupied with the concrete design of laws, while monarchs issue mere directives or royal decrees. Thus, they can employ divide-et-impera policies among their political rivals much more easily (Byman and Green 1999) andcan delegate responsibility for unpopular or failed decisions to the legislative. Because of the comparatively broad basis of their legitimacy, they are less bound to certain allied elites and can choose among different groups more freely (Anderson 2000; Ben-Dor 2000).
Also, they are less vulnerable to open contradictions than republics are. Etymologically and historically, as res publicae, the latter's legitimacy is officially derived from popular sovereignty, forcing their elites to at least nominally fight against nepotism, closed elites, and a lack of transparence and participation in the political system. De facto, of course, as authoritarian regimes, they are tolerating or even fostering these phenomena to solidify their rule. For monarchies, on the other hand, those issues are an accepted or at least tolerated part of dynastic systems, and no jarring open contradiction arises (see also R. Owen 2004, 41-44). Since decision-making processes usually do not have to be justified or openly displayed, monarchies are more independent from the populace in their decisions. They thus have fewer veto players that must be included in the decision-making process, putting them on another level of political accountability.
At the same time, monarchies are not subject to comparably strict ideologies, as republics are: “[M]onarchs can stand above tribal, religious, etlmic, and regional divisions by acting as the linchpin of the political system” (Lucas 2004, 106). Friske calls that the ability to act as a pouvoir neutre (Friske 2007, 79). They are not the leader of a particular party but rather "mediators” between different societal groups (Frisch 2011; Waterbury 1970, 267-274).
Ideology is a greater restriction for republican presidents. Republics in the region came into being mainly from social-revolutionary regime changes and, at least in the first decades of their existence, based their legitimation on ideology and revolutionary values (cf. Hinnebusch and Ehteshami 2002, 335-337). For this reason, their political maneuvering is heavily restricted, even more so on foreign policy. The basic tenants of pan-Arabism, the central revolutionary ideology of the period of decolonization, are related to foreign policy: solidarity among Arab states and hostility toward (neo-)imperialism and the “Zionist entity” Israel. These tenants form unambiguous principles that form a moral imperative and thereby considerably restrict the foreign policy choices of leaders. Open alliances between these states and their ideological enemies become almost impossible. Even informal ones are complicated, as their uncovering would have serious consequences for the legitimacy of a regime based on pan-Arabist principles. It implies both less flexibility in allowing alliances with their ideological enemies and possible chain ganging deriving from the pan-Arab solidarity (Christensen and Snyder 1990). If pan-Arabism under Nasser at first bolstered Egypt’s influence, it later became a burden to the president, who was bound by the role of the “Arab hero”, which led to confrontations with the West and Israel and eventually chain ganged him into the catastrophic 1967 war (Hinnebusch and Shama 2014, 87).
The lack of ideological rigor in Middle East monarchies leads to moderation with no clear enemies but also no clear friends: “But when the crisis passes, they once again seek some regional middle ground, avoiding, if at all possible, friendships that are too close and also enmities that are too intense” (Gause 1994, 121).
Both can serve to prevent war - because there is no clear and demonized enemy to attack and because having fewer prescribed friendship responsibilities means that it is unlikely to be dragged into a war of a "friend's” making.
The monarchies, during the period of the “Arab Cold War” on the side of the "reactionary” and “counter-revolutionary” powers, do not have those legitimacy problems to the same extent. This does not mean that they can completely disregard the opinion of their people sympathizing with the ideological currents dominating certain periods. However, there is no open contradiction between the base of their legitimacy, mostly detached from foreign policy, and their actual foreign policy and alliance behavior that does not directly threaten said legitimacy (cf. R. Owen 2004,41). This presupposes more freedom in foreign policy behavior since ideological friend-fiend schemes are less decisive for alliance or conflict. Foreign policy can thus become more "pragmatic” (cf. also: Ktihnhardt 2012).
The preference for pragmatism over ideology partially explains alliances of monarchs with Western hegemons as well as the more relaxed relationships with Israel - policies almost unthinkable for most social-revolutionary republics with strong and explicit anti-(neo)colonial rhetoric. As fervent pan-Arabism has declined significantly even in former revolutionary republics and as they have taken on other characteristics of monarchies, hybridizing from “jumhuriyya” (republic) to “jumhikiyya", a portmanteau combining the Arab terms for monarchy and republic, such alliances became more open in practically all states of the region (Ibrahim 2000). Nonetheless, the ideological underpinnings of republican rule as well as core institutional differences could not be shed. To the contrary, the transformation from socialist republics into hereditary autocracies ultimately undermined their rule because it went counter to their foundation (cf. R. Owen 2012). If ideology plays a lesser role in monarchies than in republics, it can explain both enmity and amity. Monarchies should be less prone to be dragged into wars along ideological lines.
But there are also ideological differences among monarchies. The lack of ideology is not equally spread out among all of them (the same, of course, applies to republics) - a case in point is Saudi Arabia, which has strong ideological underpinnings due to the importance of Wahhabism for the regime's legitimacy. Therefore, the pragmatism of Saudi foreign policy is restricted. First, it is one of only three monarchies that never established any kind of open relations with Israel; because of the domestic pressure, the US moved their air force base from the Prince Sultan Air Base to al-Udeid in Qatar in 2003 (Blanchard 2011; Rasheed 2007, 8, 83-95.135). Second, along with Jordan and Morocco, it is also one of the three monarchies in the region that participated in more than one war.
First indications that there might be some truth to these arguments are the remarkably good relations of the monarchies with Israel, even before the consequences of the Arab Spring solidified Iran as the common enemy. If there are significant group differences in the levels of ideology and pragmatism between monarchies and republics, they should exhibit themselves more clearly in issues that are highly contested and ideologically loaded, like relations with Israel and the Arab-Israeli conflict. The more ideologically committed, the harder it would be for Arab states to maintain cooperative relations with Israel. Indeed, on the spectrum of cooperation intensity, monarchies have consistently less-hostile and more-"normalized” relations with Israel than republics do, not just since the Arab Spring period but historically as well. Although only Jordan has a peace treaty, the monarchies often hosted trade missions and even had mutual visits (see Alpher 2015, 114; Melman 2011; Roth 2015).
Although the sheer accumulation of the monadic effect might lead to the patterns seen in dyadic interaction, we should still see an independent effect of monarchic peacefulness, even in interactions with republics. If the small difference in general bellicosity presented in the tables from earlier would prove to be a statistical artifact, there would not be much left to explain. This book therefore focuses on a dyadic explanation of joint monarchism.