Opposition inclusion and exclusion in the Arab world: Evidence from a new dataset

ABSTRACT

The literature on the Arab world is full of accounts on how authoritarian regimes manipulate the secular-Islamist cleavage to remain in power. This literature, however, has two gaps. First, there is no comparative data on the regional patterns of inclusion and exclusion that characterize regime-opposition relations in the Arab world. Second, studies usually do not account for the different shades of secular and Islamist groups. This research note bridges both gaps by introducing an original dataset: the Arab Opposition Power Relations dataset. The dataset codes power relations between four secular and Islamist groups in 13 Arab countries between 2005 and 2016 to a total of 624 group observations. Descriptive analysis of the dataset illustrates the changing patterns of inclusion and exclusion within and across countries both before and after the 2011 Arab uprisings. In so doing, the dataset opens new research frontiers for the vibrant literature on the resilience of authoritarianism in the Arab world.

KEYWORDS:

• Secularists
• Islamists
• authoritarianism
• Arab world
• dataset

Introduction

To remain in power, Arab authoritarian regimes have used a wide array of ‘authoritarian upgrading’ strategies (Heydemann, 2007). By manipulating the secular-Islamist cleavage, prevalent at both mass and elite levels (Ciftci, 2013; Wegner & Cavatorta, 2019), those regimes have attempted to ‘juggle competing social interests’ (Albrecht, 2005, p. 389).¹ In the words of Bank (2004, p. 160), the decision of regimes to include or exclude secular and Islamist opposition groups aimed at ‘integrating into [the] system of rule a broad spectrum of politically influential individuals or social groups while at the same time making sure that none of these actors develops an alternative power base that might become autonomous from the top figure’s immediate grip’.

This literature, however, has two gaps. First, we lack comparative data on the broad patterns of inclusion and exclusion that characterize regime-opposition relations in the Arab world.² Second, most accounts treat secularists and Islamists as unitary actors without accounting for their internal divisions.³ In his ‘protection racket politics’ framework, Brumberg (2002) treats secularists as one camp and Islamists as another without accounting for their within-group differences. Similarly, Lust-Okar (2004) divides the secular and Islamist opposition spectrum into radical and loyalist or legal and illegal.

This research note contributes to the literature by introducing an original dataset: the Arab Opposition Power Relations dataset.⁴ The dataset offers the first attempt to uncover the regional patterns of inclusion and exclusion of secular and Islamist groups in 13 Arab countries both before and after the Arab uprisings.⁵ I use a seven-point scale to code whether each secular and Islamist group is included or excluded from political power. In addition, the dataset reflects the intra-group divisions of both secularists and Islamists. I divide secularists into liberal and exclusionary based on Schedler & Sarsfield’s (Schedler & Sarsfield, 2007) typology. Following Bokhari and Senzai (2013), I divide Islamists into participatory and conditional.

The distinction between participatory and conditional Islamists, on the one hand, and between liberal and exclusionary secularists, on the other, better captures the distinctive ideological positions of each group, as will be elaborated in the next section. This classification also responds to scholarly calls to go beyond the radical-moderate divide in classifying ideological groups (Schwedler, 2015). Online Appendix 1 provides details on the four secular and Islamist groups, whereas Online Appendix 2 explains in detail the process of identifying political parties and organizations representing the different shades of secularists and Islamists.

The research note is structured as follows. Section 2 introduces the coding scheme of the dataset. Section 3 offers a descriptive analysis of the dataset both before and after the 2011 Arab uprisings. Section 4 concludes.

Coding scheme of the Arab power relations dataset

Table 1 presents the coding scheme of the dataset, while Table A2 includes the full coding criteria adopted. The unit of analysis is the group-country-year. The dataset uses a revised coding scheme borrowed from the widely cited Ethnic Power Relations (EPR) dataset (Vogt et al., 2015). Each Islamist or secular group, therefore, falls into one of the following seven-point categories: repressed, informal toleration (powerless), formal recognition (powerless), junior partner, senior partner, dominant and monopoly. In contrast to democratic regimes, where the opposition has a genuine chance of unseating the government, opposition groups in authoritarian regimes compete for a subtle form of political power. This takes two forms: first, influencing ‘rule making’ that is the rules governing political life and second, shaping ‘policy making’ which are mainly the policies defining the implementation of those rules (Albrecht, 2010, p. 6–7). This is how power is used in this research note.

Table 1. Coding scheme of the Arab opposition power relations dataset.

Scale	Category	Description
1	Repressed	Ideological group members are subjected to active, intentional, and targeted repression by the state, with the intent of excluding them from political power.
2	Informal toleration (powerless)	The ideological group holds no political power (or does not influence decision-making) at the national level of executive power – although without being explicitly discriminated against.
3	Formal recognition (powerless)	The ideological group holds no political power (or does not influence decision-making) at the national level of executive power but they are recognized as a legal political player in the form of a political party.
4	Junior partner	The ideological group participates as a junior partner in government.
5	Senior partner	The ideological group participates as a senior partner in a formal or informal power-sharing arrangement.
6	Dominant	The ideological group holds dominant power in the executive but there is some limited inclusion of ‘token’ members of other groups who however do not have a real influence on decision making.
7	Monopoly	The ideological group holds monopoly over power in the executive to the exclusion of members of all other groups.

Source: Author.

Note: The coding scheme of the Arab opposition power relations dataset is a revised version of the Ethnic Power Relations dataset coding (Vogt et al., 2015). The original EPR coding uses ‘discriminated’ instead of ‘repression’. Discriminated is more appropriate for examining ethnic groups but repression is clearer since it indicates the direct exclusion including via the use of force, which is not that obvious with ‘discrimination’.

In comparison to the EPR, a minor revision is made to reflect the nature of Arab politics. Powerless is divided into two sub-categories: formal recognition and informal toleration (Albrecht & Wegner, 2006). This distinction between formal recognition and informal toleration is part of the Arab regime’s menu of manipulation in terms of ‘who is or is not allowed to play the formal game’ (Lust-Okar, 2005, p. 169). Formal recognition means that a relevant ideological group is allowed to register as a legal political party (or political society in some countries) and to compete in elections. Informal toleration means that an ideological group is allowed to participate in the political field in a quasi-legal way, making them more subject to the discretion of the ruling regime. This was the case, for instance, with the Muslim Brotherhood (MB) in Egypt pre-2011 (Brownlee, 2007, p. 125).

In interacting with the secular and Islamist opposition, a regime is guided by the opposition’s strength and influence.⁶ Both are not mutually exclusive. Strength refers mainly to the ideological groups’ organizational (and electoral) strength and whether they have a large following and grassroots presence (Albrecht & Wegner, 2006). This is the case with several Islamist groups across the Arab world (Brooke, 2019). While influence, sometimes, follows from strength, this is not always the case. Most secular groups in the Arab world are influential not because of their grassroots presence but due to their access to media, including their own newspapers, and the relative freedom they enjoy in organizing public activities including electoral campaigns (Hamzawy, 2008, p. 7).

It is noteworthy to highlight that institutional power relations, that is whether groups are included or excluded from power, are exogenous to the ideological categorization of each group. While power relations reflect institutional settings, the categorization of each group signals their ideological attitude towards democracy. On the Islamist side, participatory Islamists accept the democratic rules of the game; compete in elections; form parties whenever allowed and believe that Islamic sharia and democracy are compatible. In contrast, conditional Islamists’ commitment to democracy is unclear and questionable, and democracy is only means to a greater end: implementing Islamic sharia (Bokhari & Senzai, 2013, p. 44–47). On the secular side, exclusionary secularists are in favour of excluding one or more groups, Islamists in our case, from the political field (Schedler & Sarsfield, 2007, p. 653). Some scholars go as far as to consider exclusionary secularists as liberals, but not democrats, given their exclusionary attitude towards Islamists (Shehata, 2013). Liberal secularists support the inclusion of Islamists and have frequently cooperated with them (Abdelrahman, 2009; Kraetzschmar, 2014). As a result, since the cutting line is mainly ideological, groups do not change their ideological categories.

This does not, however, contradict the inclusion-moderation thesis. According to this thesis, Islamist parties moderate their behaviour and ideology out of their inclusion in the political system, mainly through the vehicle of electoral participation and interaction with non-Islamist political actors (Tepe, 2019). Applied to the Islamist groups here, one might expect conditional Islamists to become participatory Islamists because of their inclusion and participation in elections. However, this is not necessarily the case. In the words of Schwedler (2011, p. 359), ‘changes in political opportunity structure provide the strategic logic for Islamist groups to participate, but those incentives do not always lead to ideological moderation, let alone in consistent or predictable ways’. In other words, behavioural moderation cannot be taken as a proxy or an antecedent for ideological moderation, thus leading to changing the categorization of groups. While ideological moderation happened in the past as was the case with the PJD in Morocco in the 1980s (Willis, 2004, p. 55), no instance of ideological moderation is observed in the time period covered by the dataset to warrant changing the categorization of groups.

My coding rests on secondary sources including the Economist Intelligence Unit’s quarterly reports, the CIA World Factbook, Freedom in the World reports and relevant academic and policy literature. Online Appendix 3 details the coding procedures with examples from the different cases. Online Appendix 4 provides the complete coding of the Muslim Brotherhood in Egypt between 2005–2016 as an illustrative example. Such a coding method is not unique and follows other datasets such as the Religion and Conflict in Developing Countries dataset (Vüllers et al., 2015). The collection and coding of the data, on which the analysis of this research note is based, also follows the scholarly recommendations to facilitate transparency ad replication (Salehyan, 2015).

Patterns of opposition inclusion and exclusion in the Arab world

This section offers a descriptive analysis of the dataset. Table 2 presents an overview of power relations for the four ideological groups in 13 Arab countries, comprising a total of 624 group observations. The dataset covers a twelve-year period between 2005 and 2016. This temporal focus is driven by the motivation to capture the changing power relations in the post-2011 period and to contrast that to a comparative period before the Arab uprisings. Therefore, 2016 is chosen as the endpoint at which the post-2011 regional turmoil has mostly stabilized (Lynch, 2015, p. 3).⁷ The full dataset is available in Online Appendix 5.⁸ The below overview starts with the distribution of power relations per group before explaining them across groups. Liberal secularists are repressed in one-third of the cases (33.3 per cent), formally recognized in slightly over another third (37.1 per cent) and, in a less frequency, informally tolerated (12.8 per cent). They also shared power as junior and senior partners in 7.7 per cent and 8.3 per cent of the cases respectively, while only dominating power in 0.6 per cent.

Table 2. Opposition power relations in 13 Arab countries.

	Liberal secularists	Exclusionary secularists	Participatory Islamists	Conditional Islamists		Total
Repressed	52 (33.3%)	8 (5.1%)	45 (28.8%)	19 (12.1%)		124 (19.8%)
Informal toleration (powerless)	20 (12.8%)	36 (23.1%)	28 (17.9%)	75 (48%)		159 (25.4%)
Formal recognition (powerless)	58 (37.1%)	104 (66.6%)	64 (41%)	46 (29.4%)		272 (43.5%)
Junior partner	12 (7.7%)	0	7 (4.4%)	12 (7.6%)		31 (4.9%)
Senior partner	13 (8.3%)	8 (5.1%)	11 (7%)	0		32 (5.1%)
Dominant	1 (0.6%)	0	1 (0.6%)	4 (2.5%)		6 (0.9%)
Monopoly	0	0	0	0		0

Source: Author.

Note: The raw numbers refer to group-country-year observations for each group over a twelve-year period.

Exclusionary secularists are different than their liberal counterparts due to the fact they are repressed in only 5.1 per cent of the cases. Participatory Islamists look more similar to liberal secularists than to exclusionary secularists, albeit with a less frequency of repression (28.8 per cent). Conditional Islamists are the highest group in terms of informal toleration (48 per cent) but the least to be formally recognized by the ruling regimes (29.4 per cent). They are also the second least repressed group after exclusionary secularists with only 12.1 per cent. They, however, shared power as junior partners in 12 cases (7.6 per cent) and dominated power in four cases (2.56 per cent).

To contextualize these frequencies, I provide a number of examples to illustrate the transition of one group from one category to the other. In Tunisia, liberal secularist parties namely the Congress for the Republic and the Democratic Forum for Labour and Liberties (Ettakatol) were repressed by Ben Ali’s regime between 2005 and the 2011 Arab uprisings, leading them to call ‘for the overthrow of the regime … rather than its reform’ (Storm, 2014, p. 6). After the 2011 Jasmine Revolution in Tunisia, they joined the government as senior partners along with the participatory Islamist group, Ennahda, which came to be known as the ‘troika’ (Marks, 2014, p. 10).

In Morocco between 2005–2011, the PJD, representative of participatory Islamists, was officially recognized for having their own political party and competing in elections. In contrast, the Justice and Charity movement (AWI), belonging to conditional Islamists, was informally tolerated as evidenced by their operation of over 200 social and cultural associations (Enhaili, 2010, p. 342; Willis, 2013, p. 537). Between 2012 and 2016, the PJD became a junior partner after Abdelilah Benkirane was appointed as prime minister while AWI’s status did not change (Cavatorta, 2016, p. 95). In contrast, in Kuwait, all four groups retained their formal recognition status throughout 2005–2016 by virtue of their representation in parliament as political blocks (Hafidh & Fibiger, 2019, p. 117).

In terms of the distribution of power relations across groups, formal recognition is the most frequent power relation with 43.5 per cent of the cases, followed by informal toleration with 25.4 per cent. Surprisingly, repression is the norm in only one-fifth of the cases (19.8 per cent), which lends support to the fact that Arab authoritarian regimes depend less on the exclusion of ideological groups (Cavatorta, 2007) than on their partial or full inclusion in the political scene: informal toleration or formal recognition respectively. While serving as junior and senior partners is far less frequent, they have a non-negligible share of 10 per cent with only power dominance accounting only for 0.9 per cent.

The section now moves to present changes in power relations that took place during the period under examination. As shown in Table 3, seven countries saw a change in opposition power relations while six countries saw no change. All changes took place in the post-2011 uprisings period. The countries that saw change include countries that witnessed, not only the overthrow of their long-time autocrats such as Egypt, Libya, Tunisia and Yemen, but other countries such as Morocco, Saudi Arabia and the United Arab Emirates (UAE). That the latter three are all monarchies indicates that changes in the patterns of secular-Islamist inclusion and exclusion cross-cuts the regime type, be it monarchies and republics or rentier and non-rentier states.

Table 3. Change in opposition power relations per country and per group post-2011.

Change in power relations per country post-2011	Change in power relations per group post-2011
Change: Egypt (8) Libya (4) Morocco (1) Saudi Arabia (1) Tunisia (8) UAE (4) Yemen (7) No change: Algeria Bahrain Jordan Kuwait Qatar Syria	Liberal secularists: Egypt, Libya, Tunisia, UAE, Yemen (9) Exclusionary secularists: Egypt, Libya, Tunisia, UAE, Yemen (6) Participatory Islamists: Egypt, Libya, Morocco, Saudi Arabia, Tunisia, UAE, Yemen (12) Conditional Islamists: Egypt, Libya, Tunisia, UAE, Yemen (6)
N = 7 countries with change N = 6 countries with no change	N = 33 changes in power relations for four ideological groups

Source: Author.

Note: In some countries, there was more than one instance of change in power relations for some groups in the same country during the twelve-year period under study.

In addition, changes in opposition power relations happened to all four ideological groups. In other words, none of the four ideological groups were immune to changing their power relations. Participatory Islamists saw 12 changes in their opposition power relations in all seven countries. They are followed by liberal secularists (9 changes) and exclusionary secularists and conditional Islamists (6 changes each).

While a complete analysis of the dataset for reasons of space is beyond the scope of this research note, a brief look at the patterns of inclusion and exclusion in 2010, just before the eruption of the Arab uprisings, reveals two observations. First, regimes that relied on repression, instead of the formal recognition of ideological groups, saw much more turmoil during the Arab uprisings compared to those that did not. Egypt, Libya, Syria and Tunisia belong to this group of countries. This is in line with the consolidating literature showing that repression backfires (Almeida, 2003; Hager & Krakowski, 2021).

By the same token, integration into the system via official recognition, as in Algeria, Bahrain, Kuwait, Jordan and Morocco, made secular and Islamist opposition groups more likely to call for liberalizing the regime rather than challenge the system entirely (Hamid, 2011; Moss, 2014).⁹ This is, nevertheless, not a function of the inclusion-moderation thesis as the ideological groups’ decision not to challenge the system was strategically, not ideologically, induced (Brown, 2012). Yemen is a deviant case, for that the integration of all groups did not prevent them from calling for the fall of the regime in 2011. This has roots in both the weak nature of the Yemeni state and the lack of concessions that either satisfied or divided the opposition similar to the other cases (Alley, 2013; Boukhars, 2011).

Second, regimes that informally tolerated, rather than formally recognized, ideological groups offered the latter informal ways to voice their demands as a means of containing their mounting opposition. In Qatar, Saudi Arabia and the UAE, ideological groups used petitions, majalis and diwaniyat to compensate for the lack of official political recognition (Al-Rasheed, 2010; Freer, 2018). In addition, those regimes relied on a combination of traditional and religious legitimization strategies, rentier-based appeals and targeted repression to offset any threatening political opposition (Freer, 2019; Schlumberger, 2010).

Conclusion

By introducing the Arab Opposition Power Relations dataset, this research note traced the patterns of inclusion and exclusion of secularists and Islamists in 13 Arab countries both before and after the 2011 Arab uprisings. The analysis shows how patterns of inclusion and exclusion are neither dependent on regime type be it republic or monarchy (Anderson, 1991), nor whether a country is rentier or not (Biblawi, 1987). In addition, the dataset points to several changes in power relations in Arab countries, even where regimes did not break down, further signalling the importance of studying political, in addition to regime, change (Szmolka, 2017).

While the research note only offered descriptive analysis, the dataset itself can be used in further explanatory analysis. Configurational analysis such as Qualitative Comparative Analysis (Schneider & Wagemann, 2012) or typological theory (George & Bennett, 2005) can be used to uncover the different configurations of power relations that contribute to regime survival or breakdown across the Arab world. In addition, the same methods can be used to empirically examine whether the inclusion or exclusion of secularists and Islamists facilitate or inhibit cross-ideological cooperation and induce regime support. This would enable the comparative examination of the divide-and-rule hypothesis in the Arab world (Brumberg, 2002, 2013; Lust-Okar, 2005). Furthermore, the dataset can be combined with data from the Ethnic Power Relations dataset in order to understand the similar and different patterns of including or excluding sectarian and ethnic groups in the Arab world (Belge & Karakoç, 2015). The importance of the secular-Islamist cleavage does not negate the existence of other social cleavages. The Arab world, points Sadiki (2002), is not harmonious but has linguistic, ethnic, religious and ethnic cleavages. Highlighting such patterns of inclusion and exclusion is crucial for a better understanding of how authoritarianism works. This dataset provides one step in this direction.

Acknowledgements

For their extremely helpful comments on earlier drafts of the manuscript and the dataset, I am indebted to Silvia von Steinsdorff, Claudius Wagemann, Matthijs Bogaards, Holger Albrecht, Sean Yom, Lisel Hintz, Marc Lynch, Ertuğ Tombuş, Zafer Yılmaz, Maria Abad Andrade and Daniel Hegedüs. I would like also to thank the anonymous reviewers and the journal editors for their suggestions which improved the manuscript. All errors are mine.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/13629395.2022.2085951

Additional information

Funding

The work was supported by the Elsa Neumann Scholarship (2018-2021).

Notes

1. The secular-Islamist cleavage in the Arab world, according to Ciftci (2013), revolves around the role of Islam in public life with Islamists advocating for a role for Islam in politics and secularists advocating for a separation between religion and politics.

2. One exception is Lust ‘s (2011) attempt to map the trends of inclusion and exclusion of Islamists in 42 Muslim-majority countries including several Arab countries. Nevertheless, unlike the more fine-tuned approach of power relations presented in this research note, she restricts her analysis to a binary form of inclusion-exclusion of Islamists and she focuses only on Islamists to the disregard of secularists.

3. An accurate designation of secularists would be non-Islamists. However, the distinction between secularists and Islamists adopted in this research note is necessary for analytical clarity. According to Wegner and Cavatorta (2019), this distinction differentiates Islamists who advocate for using religion to regulate public life and secularists, who despite being Muslims, aim to restrict the role of religion to the private sphere.

4. Following Albrecht (2010), p. 3), I define opposition as ‘an institution located within a political system but outside of the realm of governance that has decisive organizational capacities and engages in competitive interactions with the incumbents of a political regime based on a minimum degree of mutual acceptance’.

5. Online Appendix 1 elaborates on the rationale behind focusing on the 13 countries out of the 22 member countries of the Arab League.

6. The two criteria are adapted from Sartori ‘s (2005, p. 107–110) criteria to count relevant political parties: coalition-potential and blackmail-potential. Lust (2011) adds a third criteria which is the degree to which a regime enjoys a degree of religious legitimacy as in Saudi Arabia or Morocco.

7. The dataset focuses, therefore, on the effects of the first wave of the Arab uprisings rather than subsequent waves of protest that took place in Algeria and Sudan which are to some extent still unfinished. See Grewal (2021) for an overview.

8. The dataset and the Online Appendix are available on Harvard Dataverse via the following link: https://doi.org/10.7910/DVN/JNBWL5.

9. According to Moore-Gilbert (2016), in Bahrain, only Shiite political parties challenged the system, whereas their Sunni counterparts sided with the regime.