Abstract
This article investigates a specific type of cabinet government in the Arab Middle East and North Africa (mena): the multiparty coalition. Although mostly associated with parliamentary democratic systems, coalition governments are not uncommon in the region, comprising in fact since 1990 a sizeable proportion of the cabinets formed post-election. Drawing on novel data collated by the authors, this article offers new macro-level comparative insights into some of the key parameters of coalition governance, including their formation, composition, and durability. In doing so, the article seeks not only to document and analyse the spectrum of multiparty governance in the Arab mena but advance the development of a research agenda on the subject that, whilst sensitive to local context, engages critically with, and feeds into, the broader coalitions literature.
Introduction
In autumn 2021,1 citizens in Morocco and Iraq were called to the polls in routine elections to their national legislatures.2 As historically both polities feature highly fragmented party landscapes, it was widely anticipated that, rather than handing victory to a single political party, these elections – like the ones before – would produce results that necessitated the formation of a coalition government. And indeed, following uncharacteristically swift (by regional standards) post-election negotiations, Moroccans awoke to news that a new multiparty coalition had been formed. In Iraq, meanwhile, it took (party) political elites over a full year to negotiate a new broad-based coalition government.3
Constituting but the latest coalition events in the politics of the Arab Middle East and North Africa (henceforth Arab mena), the Moroccan and Iraqi experiences are not unique, but representative of a type of government that has become commonplace in the region since the age of political liberalisation in the early 1990s. A simple counting exercise reveals that between 1990 and 2022, a vast majority of the region’s regimes featuring party pluralism were headed at one point or another by a coalition government.4 As illustrated in Table 1, over the past three decades, no less than eleven out of the sixteen such regimes – that is nearly 70 percent – experienced between one and up to fourteen episodes of coalition governance, putting into sharp relief the need to scrutinize the trademarks and workings of this hitherto underresearched type of government in the Arab mena context.
This contribution to the special issue presents a macro-level study of coalition governance in the Arab mena, drawing on a novel set of data that comprise a total of 42 multiparty elections and 60 coalition events across eleven countries and a federated region.5 Addressing a lacuna of regional scholarship in the field, it offers a first comparative investigation into coalition governance in the region, homing in on the length of time it takes for such governments to form post election, why in some instances this process faces significant delays (bargaining delays), what size and shape (levels of political fragmentation) these governments take, how long they tend to last (coalition durability), and why some are more durable than others. Our theoretical point of departure in all this is the broader coalition literature that presents several plausible hypotheses accounting for cross, and within, country variance in coalition formation and durability. We test these hypotheses against available data from the Arab mena, focussing hereby on some of the predictors referenced in the wider literature that are internal to the coalition parties/partners themselves, and which – we assert – are hence likely to carry relevance in both democratic and non-democratic settings. On coalition formation, these include assumptions tying delays in government formation to levels of bargaining uncertainty and complexity. On coalition durability, in turn, they pertain to factors such as cabinet size and (ideological) composition, coalition strength in parliament, as well as the presence or absence of a dominant party, all of which are thought to impact the longevity of a coalition government.
Amongst others, our data reveal that it takes multiparty coalition governments in the Arab mena on average 3.5 months to form, with longer delays in coalition formation tied to the number of political parties involved, and that once up-and-running these coalitions tend to feature high levels of political fragmentation, particularly along the left-right and secular-religious cleavages. The data also reveal significant variance in the durability of such coalitions, with most of them collapsing well before their official expiry date. Notable factors found to exert an impact on coalition durability include the size of the seat majority held by a coalition government and the presence of a religious-secular cleavage amongst coalition partners. The former exerts a stabilising and the latter a destabilising effect on coalition durability.
Coalition Cabinets, Types and Change
Within the literature on authoritarian governance and regime-society relations in the Global South, the term ‘coalition’ has been deployed to denote (temporary) collaborative endeavours of varying kinds. The term is probably most widely associated with occurrences of cooperation amongst various opposition actors and groups engaged in contentious politics vis-à-vis incumbent regimes. This is the case, for instance, in some electoral studies that have used the terms ‘coalition’ and ‘alliance’ interchangeably when referencing the formation of pre-electoral pacts between two or more political parties.6 It is also referenced in social movement research on broader societal/opposition contention to describe the coming together of, at times, ideologically disparate groupings, organisations, and/or activists for the purpose of joint action on a set of specific objectives. The October 18 Collective in Tunisia and the Higher Committee for the Coordination of National Opposition Parties (hccnop) in Jordan are two examples of such cross-ideological reform coalitions in the Arab mena.7 Outside the realm of opposition politics, the term ‘coalition’ has found application in research on authoritarian governance to designate the distinctive set of traditional, bureaucratic and coercive institutions, elites and allied societal forces that underpin most authoritarian regimes and their quest for survival, particularly in the countries of the Arab mena.8
This article is not concerned with any of the above conceptualisations of a ‘coalition,’ but with one that is widely referenced in the comparative political science of Western democracies, yet not so far within the Arab mena context. What we refer to here is an understanding of the term ‘coalition’ that denotes a type of cabinet government formed jointly by two or more political parties in circumstances where a single party is unable to muster a working parliamentary majority of its own following an election, or in times of national crisis/transition. Prevalent mostly in parliamentary and semi-presidential systems, such coalitions are widely thought to carry two principal trademarks: (a) they are temporary arrangements, with the political parties involved committing to a joint programme of governance spanning the lifetime of a parliamentary term, and (b) they involve a division of cabinet portfolios and other government positions amongst its member parties (also referred to as portfolio coalitions).9
Whilst coalition governments come in all shapes and sizes, at a most rudimentary level they can be categorised based on their combined parliamentary seat shares into the following three sub-types: the surplus (also at times referenced as oversized), minimal-winning, and minority coalition.10 As the term suggests, the surplus coalition contains one or more political parties which are not strictly needed to build/sustain a simple majority in parliament and thus whose presence in cabinet is essentially dispensable. Examples of such coalitions include the broad-based national unity governments formed in Yemen (1993), Iraq (2005), and Tunisia (2011) following regime change. Minimum-winning coalitions, in turn, can be defined, according to Warwick and others, as multiparty governments that command a simple parliamentary majority but cannot ‘afford to lose any member-party’ without losing its majority.11 Minority coalitions, lastly, are multiparty governments whose member parties do not hold a majority in parliament and therefore rely on some form of ‘confidence and supply’ agreement with third parties (and/or independents) for their survival.12
Although any of these coalitions may see out their full term in office, this is not always guaranteed, which is why it is important to delineate the events that earmark coalition change; that is the termination of one coalition cabinet and the formation of the next. Following Shofield, this research counts as coalition change any event that alters the multiparty composition of a cabinet. This can entail a regularly scheduled general election, the withdrawal mid-term from government of one or more coalition partners, the merger of, or splits within, governing parties, and the fall of an entire coalition government in the wake of a political crisis and/or a parliamentary vote of no-confidence. In line with these criteria, we do not class as coalition change any cabinet reshuffles or changes in prime minister that essentially leave intact the party-political composition of the government.13
Coalition Cabinets in the Arab mena
Research on political parties in the Arab mena has traditionally been limited and linked to the paradigms of democratization and authoritarianism,14 which explains why, overwhelmingly, scholars have focussed on the study of Islamist parties and their positioning towards democratic principles. Over the past ten years, however, a noticeable change has been afoot, manifest in a substantial growth in comparative research on Arab party politics, as scholars and policymakers alike increasingly realise that the role of political parties has a profound impact on the politics of the region.15 Amongst others, this literature has shed light on themes such as how political parties organise and conduct electoral campaigns, the linkages between parties and voters,16 the electoral weakness of the political left,17 and the relationship between political parties and the media,18 precipitating a noticeable shift in focus from descriptive-historical accounts to analyses embedded in the wider theoretical literature.
This said, there remain plenty of productive avenues of inquiry to explore when it comes to party dynamics and party politics at a broader systemic level. One of these avenues concerns the study of multiparty coalitions in the Arab mena which, by moving beyond more generic debates about what kind of institutions can/ought to be associated with either authoritarian resilience or democratization, affords us a better understanding of the dynamics that shape partisan politicking and governance in the region. To be sure, some scholarship on the Arab mena exists that does explore aspects of coalition governance, mostly in the context of broader analyses of authoritarian cooptation, democratization, sectarianism, and consociationalism. In most of these studies, however, coverage of coalition cabinets tends to remain limited and – in the case of the cooptation literature – blindsighted by instances of coalition governance where authoritarian cooptation is not at stake, such as, for instance, in present-day Iraq, Iraqi Kurdistan, Lebanon, and Tunisia. For the most part, as well, this literature is lacking in any region-wide comparisons and/or linkages to the wider political science literature.19 Consequentially, little intellectual exchange has materialised between regional and global political science research on coalition governance, which is manifest in a coalition literature whose theoretical and empirical insights derive overwhelmingly from liberal democratic contexts and a nascent regional literature little engaged with wider theorising in the field.
Addressing this mutual neglect, we seek answers to the following questions: what are the principal characteristics of coalition governance in the Arab mena? How long does it take for such governments to form and what shape do they take? How (ideologically) cohesive are these coalitions and do they stay in power for the full parliamentary term? To probe these questions, we collated data on all coalition governments formed in the Arab mena between 1990 and 2022, the result of which is a dataset comprising 60 such governments in eleven Arab polities and one federated region (Iraqi Kurdistan). It holds information on coalition start-/end dates, duration (in days), periods of caretaker governance (in days), coalition size, and levels of political (ideological) fragmentation.20 In the following pages, we will first examine comparatively some key trademarks of cabinet formation, before proceeding to an analysis of levels of coalition fragmentation and some of the predictors thought to shape coalition durability in the Arab mena.
Formation
Although multiparty coalitions are not uncommon in presidential systems, – as Chaibub, Przeworski, and Saiegh have shown21 – for the most part they remain associated with parliamentary and (to a lesser extent) semi-presidential systems that feature cabinet governments led by a prime minister and responsive to parliament.22 In circumstances where elections do not produce a clear winner, both systems harbour strong institutional incentives for political parties to join forces in a coalition government that can command a parliamentary majority. This is the case also in the Arab mena, where – as highlighted in Table 2 below – between 1999 and 2022, well over two-thirds of coalition governments were formed in the region’s parliamentary systems, including in Iraq, Iraqi Kurdistan, Lebanon, Libya (2012), Morocco, and Tunisia. Although less common, coalition governments have also been established in some of the region’s semi-presidential systems, most notably in Algeria and Yemen, where coalition cabinets have become the norm, but also in Syria (before and since the civil war) and to a much lesser extent in Mauritania (2007–8), Palestine (2007), and Egypt (2012).
Research on government formation in the European context reveals that, whilst on average it takes just over a month for new governments to form following a general election, there is significant within- and cross-country variation when it comes to the duration of this process.23 Several hypotheses have been put forward to account for this variance with mixed empirical results.24 Most prominent amongst these are assumptions tying delays in government formation to levels of post-election bargaining uncertainty and complexity, or a combination thereof. Bargaining uncertainty pertains, broadly speaking, to circumstances in which the negotiating political parties hold little or imperfect information about one another’s policy/office preferences and objectives and where hence it takes time to “accumulate information about their partners’ preferences and what kind of compromise is acceptable for them.”25 Such uncertainty is thought to be particularly pronounced in the aftermath of a general election that – when precipitating changes in the party political composition of parliament as well as high rates of elite turnover – is likely to reduce the efficacy of established inter-personal relations and prior bargaining experiences for prospective negotiating partners.26 Bargaining complexity, in turn, relates to the number of political parties engaged in, and the level of (ideological) fragmentation, marking the coalition negotiations. According to Martin and Vanberg, the basic premise hereby is this: the fewer the parties involved and the more adjacent they are on policies, the faster they will be able to conclude these negotiations.27
In the Arab mena, delays in the formation of coalition governments are not uncommon. As highlighted in Table 3 below, on average it takes such governments about 3.5 months (~110 days) to form following an election, with individual coalition negotiations lasting from between less than a month to, at its extreme, well over a year. Bearing in mind the relatively small number of cases to hand (N=39),28 and the need for further in-depth scrutiny, it appears that here too factors such as bargaining uncertainty and complexity hold explanatory weight in accounting for some of the observed variance in the duration of coalition government formation. However, they do this in different ways to what has been theorised in the European context. For one, the data suggest that regional above-average delays in government formation tend to coincide with early post-transitional contexts, featuring nascent and, at times, highly volatile party systems that are likely to produce bargaining environments in which prospective partners not only know very little about one another’s policy/office preferences and objectives, but about the ins and outs of negotiating such coalitions more broadly. This was the case, for instance, in Iraq (2005; 2010), Libya (2011–12), and Tunisia (2011–12), all of which experienced the emergence of new party landscapes following the collapse of authoritarian governments, and with it comparatively long post-election coalition negotiations. Wherever such negotiations tended to be particularly crisp, as in the Algerian, Syrian, and Yemeni contexts, we suggest, in turn, that this is likely to have been facilitated by a combination of proximal and broader system-level factors conducive to speedy coalition formation. These include the presence of predominant regime parties (all three cases) that function as an anchor in coalition negotiations and significantly reduce bargaining options, the formation of pre-electoral alliances by prospective coalition partners (Syria), and the agreement by outgoing coalition partners to continue in office following an election should the parliamentary arithmetic permit it (Algeria 2007, 2017; Yemen 1997). Crucially, as highlighted in the literature on authoritarian co-optation, they involve political regimes within which the wielders of genuine policy-making power (e.g., presidents, monarchs, armed/security forces) can shape coalition formation due to the pressure they can exert on the composition of the negotiating parties and their direction.
Our data also appear to support Martin and Vanberg’s hypothesis linking delays in coalition formation to the number of political parties involved in the negotiating process.29 In our dataset the variables ‘duration of coalition formation’ and the ‘number of political parties in a coalition’ were found to be moderately positively correlated (r = .453; p = .004),30 suggesting that the larger the number of political parties involved, the longer coalition negotiations tend to drag on. The most recent coalition negotiations in Iraq following the 2021 general elections – which involved a large and diverse set of parties, pre- and post-electoral alliances, blocs, and coalitions and dragged on for over 12 months (382 days) – are a testimony to this point. Meanwhile, we find little conclusive evidence in support of Martin and Vanberg’s second hypothesis linking bargaining duration to the level of ideological distance extant between the negotiating partners. In the European context, this distance is associated most widely with policy differences on the left-right spectrum that, wherever substantial, are expected to increase the complexity of inter-party bargaining over acceptable compromises on relevant policy domains and as such the duration of the negotiation process. In the Arab mena, meanwhile, the available evidence suggests that, rather than being tied to economic distance, bargaining delays appear to be particularly pronounced wherever coalition governments are characterised by significant inter-party divisions along ascriptive lines. This transpires from Figure 1, which shows that bargaining periods well above the region-wide average (~110 days) appear to be particularly common in polities featuring highly politicised ethno-sectarian and/or tribal-familial cleavages (Iraq, Iraqi Kurdistan, Lebanon).31 It stands to reason then that coalition formation is here encumbered by bargaining complexities tied not only to the number of political parties involved and the necessity for power-sharing (Iraq, Lebanon), but to the extensive horse-trading that is known to take place between and among rival ethno-sectarian/tribal protagonists over policy, (veto) powers, influential government portfolios, and resource allocations. In fact, as media reports on government formation in Iraq and Lebanon have highlighted, in both countries coalition negotiations tend to drag on largely due to disagreements between/amongst the different sectarian parties and alliances over access to power as well as over the selection of the presidency, the prime minister, and the allocation of cabinet portfolios.32
Duration of Post-Election Coalition Formation by Coalition Size
Citation: Middle East Law and Governance 15, 3 (2023) ; 10.1163/18763375-20231393
Once sworn in, what shape do these coalition governments take? In parliamentary and semi-presidential systems, the absence of a single-party legislative majority can lead to the formation of one of four types of governments: a minority single-party government, a minority coalition, a so-called minimum-winning (mw) coalition, or a surplus coalition. Within (early) Western scholarship on coalition formation, including by Riker, Warwick, and Schofield, two hypotheses dominated research on coalition types: a first, asserting that because (office-seeking) parties are intent on maximising their power/influence in government, they prefer the formation of coalition governments involving only the minimum necessary number of political parties; and a second, bringing ideology into the equation, positing that political parties are not merely office but also policy-seeking in outlook and that they hence prefer coalition governments that comprise not only the minimum number of parties, but a minimum number of parties adjacent to one another on the policy scale (so-called minimal connected winning coalitions, mcw). Testing these hypotheses, early research in Western Europe has found that a plurality of coalition governments formed (1945–1986) were indeed either mw or mcw, with surplus coalitions and minority governments being less prevalent.33
Unlike in the Western European context, these coalition predictors sit ill at ease with data from the Arab mena, given that here mw coalitions are far less common than surplus coalitions. In the period under investigation, the latter make up well over two-thirds of all coalition governments formed, with mw coalitions and minority governments coming in a distant second. Spanning a diverse range of countries, surplus coalitions have, in fact, been the norm of multiparty governance in Algeria, Iraq, Lebanon, Mauritania, Syria, Tunisia, and Yemen. Although speculative at this point, we suggest that a need for broad-based power-sharing under consociational and quasi-consociational settings (Lebanon and Iraq), as well as under exceptional domestic circumstances, such as in post-conflict (Algeria) and transition scenarios (Mauritania, Tunisia, and Yemen), may offer some pointers as to why since the 1990s such coalitions have been far more prominent than mw coalitions in the Arab mena context. To this can be added the authoritarian nature of most political systems examined here, which too is likely to have facilitated the formation of surplus coalitions insofar as rulers might ‘encourage’ (opposition) parties to participate in a coalition government to ensure ‘blame-sharing,’ whereby participation renders parties accountable to the public and enables the ruler to deflect blame in case of a crisis. For their part, (opposition) parties might accept this game of ‘blame-sharing’ to obtain access to state resources that they can in turn use for organisational survival and to cultivate patronage networks.
Coalition Political Fragmentation
One variable internal to coalition governments widely explored, particularly in relation to questions of formation and durability, is the level of political fragmentation between partners. Entering a coalition requires parties to overcome what might be significant cleavages, as it is not always possible to set up coalitions with another party – or parties – adjacent to one’s own positions and policy preferences. As several cases in established European democracies demonstrate, coalition governments can span significant cleavages, such as, for instance, the various grand coalitions in German politics or the short-lived coalition between the 5 Star Movement and the Northern League in Italy following the 2018 elections.
To identify what can be called ‘ideological distinctiveness,’ scholars establish the core positions of parties on key policy issues, as well as their self-identification with specific groups in a polity. This is usually done through a careful codification of party manifestos, from which the exact policy positions of the parties can be derived. Such a database of codified party manifestos does not exist for political parties in the Arab mena, and we therefore identified their distinctiveness through the literature and a cursory examination of their electoral platforms, where they exist. In this process, three significant cleavages were identified in no order of importance. The first cleavage is the left-right one, which is based on the positions of parties on the economy. The left is identified, broadly speaking, with opposition to the widespread introduction of market-economic mechanisms in regulating all aspects of social and economic life. The idea is that leftist parties are in favour of redistribution through taxation, opposed to market liberalization/privatizations, and are concerned with increased integration into the world economy.34 Several Arab parties fall into this category despite the general weakness of the political left. The right is instead identified with support for further market liberalization/privatization, redistribution of wealth through market mechanisms or charity, and increased integration into the world economy. In this respect, most Islamist parties, such as the various Muslim Brotherhood affiliates, would fall into the latter category.35
The religious-secular cleavage constitutes the second divide that can been identified as of significance in Arab politics. Although the existence of this cleavage is far from unanimously recognised, particularly when it comes to the socio-political values parties promote and defend,36 we purport that “the separation between Islamists and secularists makes sense insofar as the former employ religious precepts to guide public policies while the latter wish to relegate religion to the private sphere.”37 Among voters, and in the Arab media too, the secularity or religiosity of political parties is widely referred to as a significant cleavage, particularly when it comes to policy preferences related to the right of minorities or the narratives and values that should underpin national identity.38
The final cleavage is the ethno-sectarian one. Although this does not apply to all the countries in this study, the ethnic, linguistic, sectarian, and tribal fragmentation of several Arab countries has found its way into institutional politics, with several parties clearly signalling to voters which sub-national group they intend to represent. Although the appeal to a specific ethnic or sectarian group might not be explicit in the policy preferences the parties put forth, the symbols employed, the language used, and the informal or formal resources distributed to reward a specific group can be considered a testimony to the ‘real’ nature of such parties. In addition, the institutional rules in place in some countries dictate that parties be sectarian, as they are pillars of consociational systems.
When it comes to the Arab mena, our data reveal that levels of coalition political fragmentation tend to be medium to high. This is evidenced in Figure 2, which shows that over two-thirds of all coalition governments formed in the region comprise parties spanning at least two, if not all three, of the socio-political cleavages identified above.
Overall Levels of Coalition Political Fragmentation
Citation: Middle East Law and Governance 15, 3 (2023) ; 10.1163/18763375-20231393
Note: ‘Low political fragmentation’ designates coalition governments spanning a single cleavage, ‘medium to high political fragmentation’ for those spanning any two cleavages, and ‘high political fragmentation’ for those spanning all three cleavages.A breakdown of these figures furthermore reveals that, of the three socio-political cleavages prevalent in the Arab mena context, the left-right spectrum constitutes by far the most prominent divide, spanning nearly 90 percent of the coalition governments formed since the 1990s. Several plausible explanations can be offered to account for the apparent ease with which coalitions are formed across this divide. For one, it might indicate that the majority of parties choosing to enter into a coalition with an ideological opponent on key economic policy preferences are fundamentally office-seeking or, at the very least, do not consider economic policy to be a priority. This renders compromise much easier, as a party might decide to focus its attention in coalition negotiations on other policy areas and/or desirable government portfolios. Second, it could suggest that parties are lacking in developed coherent and detailed economic policies and that, therefore, they find it easier/less arduous to negotiate with ideological rivals.39 A third explanation might reside in the recognition on the part of political parties that their economic policy preferences, usually linked to redistribution of wealth and increased public spending, are unlikely to affect this policy arena due to external and internal constraints. The difficult economic situation of several Arab countries, particularly after the 2011 Arab uprisings, has meant ‘delegating’ economic policymaking to international financial institutions, which are in a position to impose severe constraints on the autonomy of elected representatives.40 This means that the stances of political parties on the economy are somewhat irrelevant, thus rendering them more open to coalition making with ideological rivals. Along similar lines, it could be suggested that preferences on economic matters are mostly rhetorical, as all parties recognise that, ultimately, there can be very little deviation from the imperatives of the market economy. Although not to the same extent, we can see similar trends in coalitions across Europe. As economic policymaking is increasingly insulated from elected representatives due to the severity of external constraints, parties at opposing ends of the political spectrum on economic matters can find common ground to form a coalition focusing on shared interests and preferences outside of the economy.41 Finally, it could be argued that in many authoritarian contexts, it is easier to compromise on economic policy because the main decisions in this policy field remain in the hands of unelected and unaccountable wielders of power, be they the military or a monarch.
Table 4 also shows that coalitions spanning the religious-secular cleavage are reasonably popular (covering two-third of all cases), though not to the same extent as those spanning the left-right divide. Although this cleavage too can be – and has been – overcome, it is arguably more difficult to do so because, for both voters and political parties, socio-political values centred on individual rights are more salient and divisive than, say, economic issues. Surveys across the Arab world over the last decade have demonstrated quite clearly that for voters and parties, socio-economic issues are the priorities to focus on. As mentioned above, however, the restrictive international environment, and the marginal room for manoeuvre parties have when it comes to economic policymaking, renders differences on other issues, including religious identity, more electorally salient/advantageous. And indeed, it has been shown that in elections, political parties often push those religious-secular differences and that party members tend to latch on to them as a marker of diversity with consequences on how media report them and voters interpret them. The greater ideological nature of the debate on individual rights – usually pitted against group rights – divides voters and political parties to a greater extent than differences on economic policy, in part because political parties can put forth very simplistic messages when it comes to economic preferences and proposed solutions. This is not necessarily the case when it comes to the vision of society they have when it comes to individual freedoms and rights,42 making compromise more difficult. This said, it is evident from Table 4 that the secular-religious cleavage has been overcome in most coalition negotiations, suggesting that the social conservatism found in many of the region’s ‘secular’ parties renders them more palatable as coalition partners for Islamists and, vice versa, that the moderation of Islamists on some of these issues is taken seriously by prospective secular partners.43
The ethno-sectarian cleavage, lastly, appears to be at first sight the most difficult divide to overcome in the coalition politics of the Arab mena, with only a third of all coalitions spanning this divide. A closer look at the cases to hand reveals, however, that any interpretation of the data ought to account for the fact that not all societies included in the study feature clear ethno-sectarian cleavages, or if they do, that these cleavages have not become politicised within the party-political system. One of the clearest examples of this latter scenario is the Front of Socialist Forces (ffs) in Algeria that, while drawing its support almost exclusively from the Kabyle (Berber) minority, does not promote policies that simply attend to the demands of this sub-national group. When all this is considered, the picture is quite different and points to the fact that where the cleavage is present, it is usually overcome. This should not lead us to dismiss the ethno-sectarian divide in party politics, of course, although its importance should not be overstated either. The reality is that overcoming this cleavage is mandated in the consociational politics of Lebanon and the quasi-consociational arrangements in Iraq, thus necessitating political parties to collaborate across the ethno-sectarian divide if a government is to be formed at all. However, it is precisely the mandated nature of such arrangements that has led to its contestation on the part of large sectors of Lebanese and Iraqi societies, where other emerging significant cleavages – the class divide and the urban/rural one – are gaining in prominence, though as yet they have found little successful expression in established party politics.44
Durability
Coalition cabinets, like their single- and non-party counterparts, are forged to last for the duration of a full legal term, although in reality of course not all do.45 When it comes to the Arab mena, far fewer coalitions see out their full term in office than those that do. As highlighted in Figure 3, of the 54 (non-transitional) coalition cabinets that were terminated sometime between 1990 and 2022, just over a third saw out their legal term in office, with a vast majority experiencing early dissolution (at times well before) their expiry date.46 A closer look at the cases to hand reveals, in fact, that with few exceptions (Syria, Algeria) coalition failures have occurred widely across most countries of the region, with nearly half of all the coalition governments formed collapsing by around the two-year mark. With non-democratic regimes prevailing in the region, this suggests that below a veneer of authoritarian stability – exemplified mostly in the longevity of powerful presidencies or monarchies and their close ties to the coercive apparatus – executive governance remains marked by significant levels of volatility. Therefore, any presumption that coalition governance under authoritarian and semi-authoritarian tutelage ought to be characterised by greater durability than in democratic settings must be put to rest or, at the very least, seriously questioned.
Number of Full-Term Coalition Governments
Citation: Middle East Law and Governance 15, 3 (2023) ; 10.1163/18763375-20231393
Research on cabinet durability as a predictor of government stability has been a key preoccupation within the coalition literature.47 Theorising in the field revolves essentially around structural and event-based approaches or a combination thereof.48 Most structural accounts home in on predictors of durability that are internal to the coalition government itself and/or the wider (party) political system. As concerns the former, for instance, scholars have found that in the (Western) European context mw coalitions49 and those comprising fewer political parties50 and/or minimal levels of ideological diversity51 tend to last the longest. Similarly, it has been suggested that the durability of cabinets is impacted by the presence/absence of a centre/core party52 and the time it takes for their formation, with scholars presenting conflicting hypotheses on how drawn-out negotiations impact cabinet durability. Whilst for some lengthy negotiations (bargaining delays) signal heightened bargaining complexity and fractious coalition dynamics, others make precisely the opposite point, asserting that wherever coalition partners take their time to hammer out detailed agreements on policy areas, they are bound to last longer in government.53
Other structural factors external to cabinet widely cited as contributing to variance in coalition durability include the specificities of the party-political arithmetic in parliament (level of party fractionalisation) and the institutional rules regulating government formation. As concerns the former, it has been suggested that, by multiplying the number of coalitions that can be viably formed, high levels of parliamentary fractionalisation are likely to decrease a party’s commitment to a coalition, particularly in times of conflict/crisis, with options available to form alternative multiparty governments. On institutional factors, in turn, scholars have singled out rules governing investiture and/or the prime minister’s ability to call early elections as notable determinants of coalition durability.54 Event-based approaches, meanwhile, focus on possible covariates of government durability that are exogenous to a cabinet and the wider political system. Much harder to foresee in the lifespan of a government, they home in on time-dependent variables, such as economic and/or constitutional crises, and their destabilising effects on (coalition) cabinets.55
Cognisant of the novelty of this research in the Arab mena context and given a population of cases exposing significant variance in national histories and regime type – featuring fledgling democracies as well as varying types of authoritarianism – we felt it prudent to limit our analysis of coalition durability to some of the internal characteristics of cabinet government identified in the literature, including coalition type, size, and (ideological) fragmentation. All these factors constitute universal trademarks of coalition cabinets, no matter what the political system and/or circumstances, and, as such, pose little methodological challenges when being deployed as covariates across vastly different cultural/regional contexts. On this basis, we estimated a linear model that included the following explanatory variables: the number of political parties in a coalition, and the presence (or not) of a left-right, religious-secular and/or ethno-sectarian cleavage within the coalition. As an alternative to coalition type, we also included in our model the seat majority held in parliament by the coalition government, thus testing for the effects of size on durability, as well as the seat proportion held by the largest party in government. Inspired by the core party hypotheses, in this latter instance, we sought to ascertain whether the presence/absence of a pre-dominant (regime) party has any bearing on coalition durability. Lastly, we inserted government type (parliamentary vs. semi-presidential system) as a control variable in our model to estimate any confounding impact this system-level institutional factor may have on coalition durability.
Since we are dealing with panel data (multiple observations per case), we opted for a gls random-effect regression analysis as our statistical model. The results of this model are presented in table 5 below. First to note here are the two R-squares, which suggest that the model is very good at accounting for variance between cases and overall. The model itself features two statistically significant explanatory variables at the 95 percent level of confidence: a coalition’s seat majority and the presence of a religious-secular cleavage. As concerns the former, the model suggests that a more comfortable seat majority is associated with an increase in the average duration of coalition governments in the region, with each additional parliamentary seat buying a coalition 5.3 extra days in power. Our model also suggests that the presence of a religious-secular cleavage has a destabilizing effect on coalition durability, being associated with an average loss of 513 days of governing. This is nearly more than one and a half years shorter than in contexts where there is no such cleavage at play. The finding is of note, not least because it pinpoints the political salience of the religious-secular divide in regional party politics,56 but because it lends further credence to some of the key findings in the literature on cross-ideological cooperation, particularly with regards to the prevalence of serious obstacles to high-level cooperation between the two camps (including e.g. legacies of mutual-suspicion/acrimonious relations and policy red-lines).57
As is evident from the wider literature in the field, explaining the durability of government is a complicated matter. Therefore, it is important to re-emphasize that this analysis constitutes a mere first (and most likely incomplete) attempt to systematically explore cross-regional variance in coalition durability. Indeed, additional modelisation (a panel framework, for example) would be needed, alongside further in-depth case-based research, to ascertain more fully the political dynamics that drive coalition survival and termination in the Arab mena.
Conclusion
Over the past decade, a growing number of scholars of Arab politics have sought to connect research on the region with the broader political science literature, thus moving beyond a legacy of scholarship on the region that for the most part explored socio-political phenomena through the prism of ‘exceptionality.’58 Acknowledging that there are specific tendencies that can be labelled as ‘regional’ and hence as incomparable, this new literature works dialectically, bringing together local case-knowledge and the wider theoretical and methodological tools of comparative politics to better grasp regional politics and their positionality vis-à-vis the global context. In so doing, it shines light on the workings of institutions, organizations and realities (e.g., courts, parliaments, coalition governments, political parties) which might otherwise be discarded as inconsequential due to the peculiarities of authoritarian governance, cultural context, or both.
The study of coalition governance in the Arab mena is a case in point. Hitherto little problematized, in this article we demonstrate not only that this type of government has become commonplace in the region – thus warranting investigation – but that there is investigative value in a deepening conversation between area studies and the wider coalition literature for the purpose of better understanding this phenomenon. Such a conversation is already underway in the fields of Arab electoral and party politics with intriguing results59 and it is hoped that this article and the wider special issue, will set the stage for a similarly enriching trajectory of research into the region’s coalition governments.
For the purpose of open access, Hendrik Kraetzschmar has applied a Creative Commons Attribution (cc by) licence to any Author Accepted Manuscript version arising from this submission. The data associated with this article are openly available from the University of Leeds Data Repository. https://doi.org/10.5518/1291.
Francesco Cavatorta’s work for this article was supported by a grant (number 435-2020-0539) from the Social Sciences and Humanities Research Council of Canada (sshrc).
Souhail Karam, “Morocco names post-Islamist cabinet as it plans economic revamp,” Bloomberg UK (7 October 2021), available at: www.bloomberg.com/news/articles/2021-10-07/morocco-names-post-islamist-cabinet-as-it-plans-economic-revamp?leadSource=uverify%20wall; Anon., “Iraq’s parliament approves new government,” Aljazeera (27 October 2022), Available at: www.aljazeera.com/news/2022/10/27/iraq-lawmakers-approve-govt-of-prime-minister-designate-sudani.
The year 1990 was chosen as a starting point for this investigation because it demarcates – within the context of broader political liberalization – the onset of a noticeable pluralization of the party landscapes in countries across the Arab mena (and elsewhere) and, with it, the emergence of coalition politics.
The authors thank Isabelle Tremblay-Brousseau and Aisha Al-Jubouri for their invaluable support in compiling the coalitions database. We also thank Prof Marc-André Bodet from Laval University, Quebec, Canada, for his advice and input in the analysis of the data.
Denis Kadima, “The Study of Party Coalitions in Africa: Importance, Scope, Theory and Research Methodology,” in The Politics of Party Coalitions in Africa, eds. Denis Kadima and H E Cassam Uteem (Houghton: Konrad Adenauer Stiftung, 2006): 10.
About societal/oppositional coalitions see e.g., Yasmine Berriane and Marie Duboc, “Allying beyond Societal Divides: An Introduction to Contentious Politics and Coalitions in the Middle East and North Africa,” Mediterranean Politics 24.4 (2019): 399–419; Hendrik Kraetzschmar, “Mapping Opposition Cooperation in the Arab World: From Single-Issue Coalitions to Transnational Networks,” British Journal of Middle Eastern Studies 38.3 (2011): 287–302.
Matt Buehler and Mehdi Ayari, “The Autocrat’s Advisors: Opening the Black Box of Ruling Coalitions in Tunisia’s Authoritarian Regime,” Political Research Quarterly 71.2 (2017): 330–346; Sean Yom, “Oil Coalitions and Regime Durability: The Origins and Persistence of Popular Rentierism in Kuwait,” Studies in Comparative International Development 46.2 (2011): 217–41.
Following Shofield, Lijphart and Dodds, the coalitions database for this research only counts political parties as members of a coalition government if they are represented in cabinet with at least one ministerial post. Political parties tolerating a coalition government from outside cabinet are hence not deemed formally part of a coalition government and will not be accounted for in the database. See Norman Schofield, “Stability of Coalition Governments in Western Europe: 1945–1986,” European Journal of Political Economy 3.4 (1987): 559; Arendt Lijphart, “Power-Sharing versus Majority Rule: Patterns of Cabinet Formation in Twenty Democracies,” Government and Opposition 16:4 (1981): 409; Lawrence C. Dodd, Coalitions in Parliamentary Government (Princeton: Princeton University Press, 1976).
Norman Schofield, “Stability of Coalition Governments in Western Europe: 1945–1986,” European Journal of Political Economy 3.4 (1987): 559.
Paul Warwick, “The Durability of Coalition Governments in Parliamentary Democracies,” Comparative Political Studies 11.4 (1979): 470.
Shofield, “Stability of Coalition Governments,” 558–559.
Algeria presents several illustrative examples of cabinet change falling short of coalition change, as defined above. During the 2002–2007 electoral cycle, for instance, the country was governed by a tripartite coalition government, comprising the fln, rnd and msp. Whilst surviving the full parliamentary term, the coalition experienced several cabinet reshuffles, two of which involved the replacement of the Prime Minister: Benflis iii (June 2002-May 2003), Ouyahia iii (May 2003-April 2004), Ouyahia vi (April 2004-May 2005), Ouyahia V (May2005-May2006), and Belkhadem I (May 2006-June 207).
Kay Lawson and Saad Eddin Ibrahim, Political Parties and Democracy: The Arab World (New York, NY: Praeger, 2010); Lise Storm, Party Politics and the Prospects for Democracy in North Africa (Boulder, CO.: Lynne Rienner, 2013).
Valeria Resta, “The Effect of Electoral Autocracy in Egypt’s Failed Transition: A Party Politics Perspective,” Italian Political Science Review/Rivista Italiana di Scienza Politica 49.2 (2019): 157–173; Raymond Hinnebusch, “Political Parties in mena: Their Functions and Development,” British Journal of Middle Eastern Studies 44.2 (2017): 159–75; Francesco Cavatorta and Lise Storm, Political Parties in the Arab World: Continuity and Change (Edinburgh: Edinburgh University Press, 2018).
Eva Wegner and Francesco Cavatorta, “Revisiting the Islamist-Secular divide: Parties and Voters in the Arab World,” International Political Science Review 40.4 (2019): 558–575.
Valeria Resta, “Leftist Parties in the Arab Region before and after the Arab Uprisings: Unrequited Love?” in Political Parties in the Arab World: Continuity and Change, eds. Francesco Cavatorta and Lise Storm (Edinburgh: Edinburgh University Press, 2018); Idriss Jebari, “The Rise and Fall of the Arab Left” in Routledge Handbook on Political Parties in the Middle East and North Africa, eds. Francesco Cavatorta, Lise Storm and Valeria Resta (London: Routledge, 2020).
Francesco Cavatorta and Nidhal Mekki, “How can we agree on anything in this environment? Tunisian Media, Transition and Elite compromises: A View from Parliament,” International Journal of Press/Politics 26.4 (2021): 822–841.
Inmaculada Szmolka, “Party System Fragmentation in Morocco,” The Journal of North African Studies, 15.1 (2010): 13–37; Harith Hasan, “Shi’a Islamist Parties in Iraq: From Opposition to Governance,” in Routledge Handbook on Political Parties in the Middle East and North Africa, eds. Francesco Cavatorta, Lise Storm and Valeria Resta (London: Routledge, 2020).
On coalition size, the data contains information on the names/numbers of alliances and political parties involved in a coalition, the parliamentary seat share held by a coalition government as well as the seat shares of the largest and smallest coalition parties.
Research by Cheibub, Przeworski and Saiegh reveals, for instance, that between 1946–1999 coalition governments, whilst more frequent in democratic parliamentary systems, have also occurred widely under presidentialism. Forged either pre- or post-election, these governments have taken the shape of both minority and majority presidential coalitions. See José Antonio Cheibub, Adam Przeworski, and Sebastian M. Saiegh, “Government Coalitions and Legislative Success under Presidentialism and Parliamentarism,” British Journal of Political Science 34.4 (2004): 565–587.
A key institutional trademark common to all types of semi-presidential systems is the presence of a dual executive comprising a popularly elected president as well as a cabinet government headed by a prime minister and responsive to parliament. See: Robert Elgie, “Semi-Presidentialism: Concepts, Consequences and Contesting Explanations,” Political Studies Review 2 (2004): 317.
Alejandro Ecker and Thomas M. Meyer, “The Duration of Government Formation Processes in Europe,” Research & Politics 2.4 (2015): 3–4.
Ibid., 3–8.
Ibid., 2.
Daniel Diermeier and Peter Van Roozendaal, “The Duration of Cabinet Formation Processes in Western Multi-Party Democracies,” British Journal of Political Science 28.4 (1998): 620.
Lanny W. Martin and Georg Vanberg, “Wasting Time? The Impact of Ideology and Size on Delay in Coalition Formation,” British Journal of Political Science 33.2 (2003): 325–27.
In the Arab mena context, it proved difficult to obtain precise data on the termination and start dates of coalition governments in-between elections, which is why in our analysis we opted to focus exclusively on the duration of coalition formation post-election, for which data is readily available.
Martin and Vanberg, “Wasting time?,” 327.
This statistic reports a two-tailed bivariate Pearson correlation. Results are similar and statistically significant when performed with a Kendall tau_b and the Spearman’s rho coefficients.
Dai Yamao, “Sectarianism Twisted: Changing Cleavages in the Elections of Post-War Iraq,” Arab Studies Quarterly 34.1 (2012): 27–51; Bassel F. Salloukh, Rabie Barakat, Jinan S. Al-Habbal, Lara W. Khattab and Shoghig Mikaelian, The Politics of Sectarianism in Postwar Lebanon (London: Pluto Press, 2015), 88–107.
See e.g., contribution by Lise Storm and Dylan O’Driscoll to this Special Issue. See also Hadad Hamzeh, “Path to Government Formation in Iraq,” Konrad Adenauer Foundation (January 2022). Available at: www.kas.de/documents/266761/0/Hamzeh+Hadad+-+Path+to+Government+Formation+in+Iraq+2022.pdf/7040ae0a-855f-8578-2621-853a264ddcef?version=1.0&t=1641553336368; Will Todman, “Lebanon’s New Government,” Centre for Strategic and International Studies (10 September 2021). Available at: www.csis.org/analysis/lebanons-new-government#:~:text=Lebanon%E2%80%99s%20New%20Government%20September%2010%2C%202021%20After%2013,Michel%20Aoun%20and%20Speaker%20of%20Parliament%20Nabih%20Berri.
Shofield, “Stability of Coalition Governments,” 560.
Eva Wegner and Francesco Cavatorta, “Revisiting the Islamist-Secular Divide: Parties and Voters in the Arab World,” International Political Science Review 40.4 (2019): 558–575.
Joseph Daher, “Hezbollah, Neoliberalism and Political Economy,” Politics & Religion 13.4 (2020): 719–747; Maryam Ben Salem, “God Loves the Rich. The Economic Policy of Ennahda: Liberalism in the Service of Social Solidarity,” Politics & Religion 13.4 (2020): 695–718.
Brandon Gorman, “The Myth of Secular-Islamist Divide in the Muslim World: Evidence from Tunisia,” Current Sociology 66.1 (2018): 145–164; Hendrik Kraetzschmar and Alam Saleh, “The Struggle for Power and the ‘Secular-Islamist’ Binary in Post-Mubarak Egypt,” in Islamists and the Politics of the Arab Uprisings: Governance, Pluralisation and Contention, eds. Hendrik Kraetzschmar and Paola Rivetti (Edinburgh: Edinburgh University Press, 2018).
Wegner and Cavatorta, “Revisiting the Islamist-Secular Divide,” 558–575.
Valeria Resta, “The Terminal: Political Parties and Identity issues in the Arab World,” in Routledge Handbook on Political Parties in the Middle East and North Africa, eds. Francesco Cavatorta, Lise Storm and Valeria Resta (London: Routledge, 2020).
The absence of fully developed party programmes/manifestos, including on economic policy, has been noted widely in the literature on mena party politics. See e.g., Marwan Muasher, “The Path to Sustainable Parties in the Arab World,” Policy Outlook, Carnegie Endowment for International Peace (November 2013). Available at: https://carnegieendowment.org/files/sustainable_arab_polit_parties.pdf.
Adam Hanieh, “Shifting Priorities or Business as usual? Continuity and Change in the post-2011 imf and World Bank Engagement in Tunisia, Morocco and Egypt,” British Journal of Middle Eastern Studies 42.1 (2015): 119–134.
Jonathan Hopkin, Anti-System Politics: The Crisis of Market Liberalism in Rich Democracies (Oxford: Oxford University Press, 2020).
Andrea Teti, Pamela Abbott, and Francesco Cavatorta, “Do Arabs Really want Democracy? Evidence from Four Countries,” Democratization 26.4 (2019): 645–665.
Kraetzschmar and Saleh, “The Struggle for Power and the ‘Secular-Islamist’ Binary.”
Irene Costantini, “The Iraqi Protest Movement: Social Mobilization amidst Violence and Instability,” British Journal of Middle Eastern Studies 48.5 (2020): 832–849.
On evidence from Western Europe see e.g., Klaus von Beyme, Political Parties in Western Democracies (New York: St Martin’s Press, 1985).
Cabinet duration is widely computed in either days or months from the day the government is sworn in until its termination.
The hazard approach by King et al., for instance, presents a theoretical model that combines events-based and structural approaches. See: Gary King, James E. Alt, Nancy E. Burns and Michael Laver, “A Unified Model of Cabinet Dissolution in Parliamentary Democracies,” American Journal of Political Science, 34 (1990): 846–71.
Lawrence C. Dodd, “Party Coalitions in Multiparty Parliaments: A Game Theoretic Analysis,” American Political Science Review, 68 (1974): 1093–1117.
Michael Taylor and Valentine Herman, “Party Systems and Government Stability,” American Political Science Review 65 (1971): 28–37.
Paul Warwick, “Ideological Diversity and Government Survival in Western Democracies,” Comparative Political Studies, 25 (1992): 332–61.
Norman Schofield, Bernard Grofman and Scott L. Feld, “The Core and Stability of Group Choice in Spatial Voting Games,” American Political Science Review, 82 (1988): 196–211.
Michael Laver and Norman Schofield, Multiparty Government: The Politics of Coalition in Europe (Ann Arbor: University of Michigan Press, 1998); Kaare Strom, “Contending Models of Cabinet Stability,” American Political Science Review, 82 (1988): 923–41.
Nathan S. Balke, “The Timing of Parliamentary Elections,” (unpublished manuscript, Southern Methodist University, 1988); Strom, “Contending Models of Cabinet Stability”; Taylor and Herman, “Party Systems and Government Stability.”
Eric C. Browne, John P. Frendreis and D. W. Gleiber, “The Process of Cabinet Dissolution: An Exponential Model of Duration and Stability in Western Democracies,” American Journal of Political Science, 30 (1986): 628–50.
This is an observation made also by Abdullah Aydogan, whose research on party system cleavages in the Arab world reveals that the religious-secular dimension remains highly important. See: Abdullah Aydogan, “Party Systems and Ideological Cleavages in the Middle East and North Africa,” Party Politics 27.4 (2021): 814–826.
Janine A. Clark, “The Conditions of Islamist Moderation: Unpacking Cross-ideological Cooperation in Jordan,” International Journal of Middle East Studies 38.4 (2006): 539–560; Jillian Schwedler and Janine A. Clark, “Islamist-Leftist Cooperation in the Arab World,” Isim Review 18.1 (2006): 10–11.
Marc Lynch, “Taking Stock of mena Political Science after the Uprisings,” Mediterranean Politics 26:5 (2021): 682–695; Francesco Cavatorta, “Overcoming Exceptionalism – Party Politics and Voting Behavior in the Middle East and North Africa,” in Larbi Sadiki (ed.) Handbook of Middle East Politics: Interdisciplinary Inscriptions(London: Routledge, 2019); Jillian Schwedler, “Comparative Politics and the Arab Uprisings,” Middle East Law and Governance 7.1 (2015): 141–152.
Francesco Cavatorta and Lise Storm, Political Parties in the Arab World (Edinburgh: Edinburgh University Press, 2018); Aydogan, “Party Systems and Ideological Cleavages”; Carolina De Miguel, Amaney Jamal and Mark Tessler, “Elections in the Arab world: Why do Citizens Turn Out?,” Comparative Political Studies 48.11 (2015): 1355–1388.
