From Peak to Trough: Decline of the Algerian “Islamist Vote”

What are the factors that facilitate or hinder Islamic political parties’ performance in elections in the Middle East and North Africa? Why did Algerian Islamists as an electoral force declined steadily over the past two decades? Why didn’t Algerian electoral Islamists present the same mobilization capacity as their counterparts in neighboring countries did in early 2010s following the Arab Spring? In analyzing the evolution of three related variables: incumbents’ power structure and political openness; electoral Islamists’ inclusiveness and unity; and the framing process of Islamic political parties to build a legitimacy, the article tries to address the questions and contribute to the theoretical framework of the political process model by applying it to a case that is typical in MENA.


Introduction
Understanding Islamic political parties 1 becomes an urgent concern following the Arab Spring, as the anti-authoritarian protests resulted in the rise of Islamists at the ballot box in 2011 in lots of countries in MENA (Middle East and North Africa) including Egypt, Tunisia and Morocco. A wider audience are now interested in learning about the motives, strength, ideologies and activities of those religious organizations seeking to ascend power by winning elections in MENA.
Existing scholarship on Islamic political parties in the MENA has paid enormous attention to explore the big topics of moderation/radicalization and participation/ non-participation. The first topic has been discussed a lot by Schwedler (2006), Driessen (2012), Guazzone (2013), Pevná (2014), Wickham (2015), and Ashour (2009), who put forward various explanatory frameworks to answer the questions of whether Islamic political parties' ideologies are based upon literal readings of Quran, Hadith and Sharia or are adaptable according to current realities. Similar emphasis has been put on religious parties' decision-making as to whether or not they participate in elections and the factors that determinate their decisions (Kepel 1994;Bukay 2007;Lewis 1996;Zeghal 2008;Sinno and Khanani 2009;Wegner 2011).
Yet, less attention has been paid to examine the mobilization capacity of Islamic political parties when they are both allowed to and are willing to run for elections. Even though some efforts have been made to interpret the rise of one particular party such as the Turkish Welfare Party (Onis 1997;Akinci 1999;Gülalp 2001), Turkish Justice and Development Party (Sezer 2002), Egyptian Freedom and Justice Party (Farag 2012;Elsayyad and Hanafy 2013), Algerian FIS (Chhibber 1996), etc., there is a lack of comprehensive theoretical frameworks which could be generalized to more cases. Among the descriptive scholarships that focus on specific parties, one common perspective is that socio-economic grievances drive Islamic political parties to popularity while lack of grievance hinders their growth (Gülalp 2001;Layachi 2000;Garcia-Rivero and Kotzé 2007;Chhibber 1996). The assumption is that Islamic political parties are called for as an alternative when the there is a wide spread discontent about unemployment, housing shortages, poverty and other problems which could not be resolved efficiently by the regime. The grievance perspective, though offers partial answers, does not offer wholly satisfactory explanations. The assumption is problematic as it could not interpret why Islamic political parties were defeated crushingly at the ballot box in Algeria, but obtained a landslide victory in Tunisia, Egypt and Morocco in the early 2010s (see Table 3) (Zoubir and Aghrout 2012), considering that the four countries were almost in the same socio-economic predicaments at the time (see Table 1).
Thus, to better understand the determinants of Islamic political parties' mobilization capacity, other explanatory frameworks are needed. In this article, I argue that the political process model is such a theoretical perspective that could offer more comprehensive interpretations.
2 Applying the political process model to MENA cases The political process model has emerged in response to the shortcomings of what Ismail (2000) calls the ''psychosocial model'' that emphasized on the socioeconomic factors and peoples psychological discontents behind social movements. Espoused by social scientists including McAdam, Tarrow and McCarthy, the political process model focuses on how political environments create opportunities that are utilized by social movement entrepreneurs through structure building and framing strategies (McAdam 1982;McAdam et al. 1996;Tarrow 1994;Kriesi et al. 1992). It contends that socio-economic grievances could only be transformed into social movements when favorable political conditions exist (McAdam 1982) (Lucardie 2000;Tarrow 1998). The model further demonstrates that how movement entrepreneurs involve their organizational structures and normative discourses to their advantage could affect the mobilization outcome (Tarrow 1994;Lucardie 2000;Jenkins and Perrow 1977).
As scholarships of political process model have mostly been developed by analyzing cases from Western developed countries, there are a couple of limitations when the framework is used to investigate the social mobilization in developing countries including the ones in MENA. For one thing, analysts of social movements, particularly Islamist movements, in MENA examine anti-government activities like violence (Hafez 2004;Alimi 2006). For another, more attention has been paid to informal political participation of movement entrepreneurs such as kinship network (Wickham 2004). Yet, it is a common phenomenon in MENA that Islamist movement entrepreneurs seek to use ''conventional means, in the form of a political party, to mobilize against a civil state'' (Eligür 2010) and co-opt with the regime in return for institutionalized participation.
Eligür has broken through these limitations by offering a political process structure to analyze the mobilization of Turkish Islamic political parties. But just as she mentioned at the beginning of The Mobilization of Political Islam in Turkey, ''Turkey is the only muslim secular democratic state'' (Eligür 2010), her framework does not necessarily fit into the other cases in MENA where the majority of Chin. Polit. Sci. Rev. (2016) 1:717-735 719 countries implement the electoral authoritarian systems which combine ''democratic and authoritarian elements'' to ''capitalize on electoral legitimacy while remaining in power'' (Diamond 2002). This article aims to step further by applying the political process approach to Algeria where a typical electoral authoritarian system has been in place since the political liberalization reform in 1989. I argue that under the electoral authoritarian institution, two opportunities are key factors that drive Islamic political parties to the forefront of the political scene: divided political environment and controlled inclusion. I further contend that an inclusive structure maintained by charismabased unity and successful framing tactics to build legitimate images are complementary elements that determine mobilization capacity. Also, it is worth noticing that the three components of the political process model are dynamically interacted with each other. In many cases, the ruling elites who largely decide Islamic political parties' opportunities try to interfere with these parties' organizational structure and framing process to further narrow their opportunities.
In electoral authoritarian systems in MENA, the power structure can basically be dichotomized into undivided environments and divided ones. In the context where ruling elites manage to reduce disputes either by finding a common ground for all factions or by establishing a hegemonic system with one very powerful faction, the regime could gather all its energy and resources to suppress the threat from all kinds of Islamist forces. Therefore, Islamic political parties are very likely to be excluded from the political scene.
By comparison, in divided political environments where little common ground is reached between different factions and none of them is at the hegemony, some incumbents tend to ''increase political openness to some political opponents while excluding others'' (Lust-Okar 2004) so as to counterbalance their adversaries within the elite circle. As for religious forces such as Islamic political parties, such openness could be in the form of either ''informal toleration'' (Albrecht and Wegner 2006) or controlled inclusion. In the case of Tunisia in 1989, Ben Ali rejected to offer Ennahda legal recognition as a political party but allowed it to participate in the election that year as independents. This offers a typical example of the regime's informal toleration of Islamic political parties. Regarding controlled inclusion, Algerian Islamic political parties' experience in the past two decades is a case in point, as the regime designed various methods to restrict their activity spaces. The most straightforward approach is to enact laws to the disadvantage of legal Islamist activists such as forcing them to remove the Islamic elements from the parties' titles and platforms. Such methods are termed as ''legal forms'' of interference by Willis (2002). Meanwhile, covert means can be used as ''supplementary measures'' (Willis 2002). One frequently used clandestine method by Algerian incumbents is to create or catalyze inner splintering within Islamic political parties.
Apart from favorable political opportunities, organizational structure is another key factor that influences Islamic political parties' mobilization effects. Compared with secular opposition parties, Islamic political parties' connections to religious facilities and organizations yield important strength for them. And that is an important reason why Islamic political parties always demonstrate a larger mobilization capacity than secular opposition parties. Nonetheless, not all Islamic political parties are able to optimize their resources. A common characteristic of the populist Islamic political parties is their success in forming a ''melting pot'' that gathers diverse sections of the Islamist movements. Exclusive Islamic political parties either fail to incorporate Islamists from different background or expel members who hold different opinions from the leadership, which makes the parties more cliquish. It is worth noticing that due to a lack of democratic culture within most of the Islamic political parties, charismatic leadership is always the most effective system to maintain the parties' internal inclusiveness and unity.
Framing process is the third critical dimension for mobilization. David and Benford put forward three major framing missions for movement entrepreneurs: identification of social problems; indication of solutions; and motivational frames that encourage the public to become active participants (Snow and Benford 1988) by using identity issues or cultural symbols to arouse sympathy (Snow and Benford 2000).
In many cases in MENA, opposition parties' campaigning platforms hardly distinguish from one to another. Whereas all opposition parties criticize corruption, housing shortages, unemployment and government incompetency, they lack comprehensive reform schemes themselves and propose nothing but ''sterile'' and ''theoretic'' plans (Adel 2012;Parks 2013). Hence, the motivational frame is the most significant mission that determines Islamic political parties' mobilization in MENA. And I argue that the most successful motivational frames in the MENA cases are the ones that could help the parties build a legitimate image. While a large population wish to see a change from socio-economic problems, an anti-regime political force need to clarify why the party itself is the most legitimate force to substitute the current incumbents, and why it is the most likely party to change the status quo.
Algeria provides a rich case study of how the above-mentioned variables facilitate or hinder Islamists' efforts to expand their support base. The Algerian case is of particular interest because the country's ''Islamist vote'' experienced a dramatic plummet in the past two and a half decades. Both the spectacular rise of FIS in the early 1990s and collapse of Green Algerian Alliance (Alliance de l'Algérie, AAV) AAV in the election of 2012 (see Table 2) stunned the Western media. The minimal vote share gained by AAV-the first Islamic electoral alliance every created in the countrys history-is of more interest when taking into account the victories of Islamic political parties in almost all of Algeria's neighbors in the previous year. This article attempts to present a three-stage political process framework and applies it to the empirical evidence of Algeria. The aim is to answer the following questions: Why do some Islamic political parties demonstrate stronger electoral mobilization capacity than the others? Why are certain Islamic political parties more successful in enlarging their support base in some periods than in others?

Emergence of Islamic political parties in the Algerian electoral scene
The contemporary Islamic political parties in Algeria were forged in the late 1980s and early 1990s after President Chadli Bendjedid initiated a political liberalization reform. Before that, Islamists only created small cultural circles around prestigious reformist ulemas (Islamic scholars) epitomized by Sheikh Abdellatif Soltani, Sheikh Ahmed Sahnoun, and Malek Bennabi, and later around their disciples-Abbasi Madani, Mahfoud Nahnah and Abdallah Djabllah. Such circles were mainly operated in certain districts of Algiers, Blida and Constantine, and received support from little more than a small population. Some clandestine armed Islamist groups did emerge in the 1970s epitomized by Mustapha Bouyali's Algerian Armed Islamic Movement (Mouvement Algérien Islamique Armée, MAIA), but again, it failed to mobilize much sympathy and did not exist long before being crashed by the regime.
In February 1989, a new constitution was announced. The most important changes it brought for the country's political life were to legalize non-governmental political associations. The permission for citizens to create ''associations with political character'' in Article 40 was further intensified by the party law promulgated in July which allowed the existence of opposition political parties, marking an end to the one-party rule in Algeria since its independence.
The revision of constitution was both a result of mass protests against dictatorship, corruption and economic decay since October 1988, and of a power struggle between the leftists and reformists represented by the President Chadli Bendjedid. Known as a reformist seeking to substitute a liberalized economic program for the socialist orientation of the past two decades, Chadli faced huge obstacle since he took power in 1979 from the leftist elites who adhered to the socialist programs implemented by his predecessor, Houari Boumediène. To counterbalance the influence of the leftists both in the state and the society, Chadli took a more tolerant position towards the Islamist movement than the two ex-Presidents (Barka 2012) and finally institutionalized the Islamists to the political stage by legalizing Islamic political parties, despite the constitution's prohibition of political parties with a religious basis (Cheriet 1992).
Benefiting from the political transform and incumbent divisions, a number of Islamic political parties emerged and expanded. Among them, the most influential parties were FIS, Hamas (HMS), and Islamic Renaissance Movement (Movement de la Renaissance Islamique, MRI). The FIS was the first Islamic political party which was founded in a mosque in Algiers called Ibn Badis right after the implementation of the constitution (Mortimer 1991), and was legalized in September. HMS and MRI were formed later than FIS and obtained legal recognition in 1990 and 1991.
When first created, the three Islamic political parties shared much in common regarding their organizational structures. To begin with, all three parties were composed of the Congress, Majlis Choura, Political Bureau, etc., and had institutes both at the national and local level. Secondly, the internal system of FIS, HMS and MRI was based on a charismatic relationship between the leader(s) and followers. Apart from a proficiency in Standard Arabic and profound knowledge of Islam that the heads of the three parties shared in common, each leader had their unique profile on which their popularity was built. For example, Nahnah, head of HMS, was renowned for both his active role in the early Algerian Islamist movement and his close ties to the Egyptian Muslim Brotherhood (Walsh 2007). The MRI's founder, Djballah, was pioneer ulema of the younger generation in the East and gained a strong influence since the 1970s in Constantine.
As for the FIS, while President Madani won a prestige from his participation in the FLN to fight for the Algerian War of Independence in the 1950s and his experience of imprisonment by the French, his deputy, Ali Belhadj, was popular for his young age, great courage and hard-line rhetoric. Such a loyalty based organization system was a source of the groups' internal cohesion, which is instrumental to keep divergences at bay.
To certain extent, all three parties created an all welcome atmosphere to Islamists from different sections. Among them, FIS incorporated the largest variety of Islamists as it gathered together all tendencies including both the moderates and radicals: Salafiyya, 2 Jazairiyya, 3 Badisiyya 4 and Afghaniyya. 5 By comparison, the core membership of HMS and MRI adopted more elitist strategies and implemented moderate programs (Willis 1996), which made them less likely to attract some of the radical tendencies. MRI's mobilization capacity was even more restricted by its concentration in the east and lack of resources to establish branch offices in every wilaya like what FIS and HMS did.
With regard to the discourses, all three parties sought to derive their legitimacy from religion and theology by referring to sections from Quran or Hadith in their platforms, calling for an Islamic system based on the Sharia, and criticizing the government policies for their deflection from Islam. More importantly, the party leaders' readings of religious texts at the very beginning were quite literal and doctrinal, leading them to take a hostile posture on any elements imported from the West including freedom of speech, multi-party politics and electoral competition. For instance, Nahnah and Madani agreed that political plurality and individual freedoms should not go beyond the limits of the Sharia. Djaballah considered some parties' idea of secularity to be illegitimate (Cheriet 1992).
Yet, FIS distinguished itself from the other Islamic political parties by claiming itself as the heir of the Algerian anti-colonial revolution. In view of the popular nostalgia for the ''good old society'' in the revolutionary era, as well as the wide spread disappointment for the regime's corruption, nepotism and incompetence, FIS used revolutionary symbols to show that they were the legitimate heirs that really inherited the core spirit of the Algerian revolution of November 1 1954-''Islam is my religion, Algeria is my nation, and Arabic is my language''. 6 The FIS combined the religious and revolutionary ideas in its discourse by declaring that the revolutionary spirits required all Algerians to be unified by their Arab and Muslim identity, and the reason why the country suffered from so many socio-economic problems was that the ruling FLN did not insist on their identity in the postindependent era and imported lots of Western values (Verges Verges 1997;Addi et al. 1992). Lacking Madani-type figures that were veterans in the revolution, neither HMS nor MRI was entitled to build a similar revolutionary discourse.
Thus, with more inclusiveness and a legitimate image based on inheritance of revolution, FIS became the largest beneficiary from the favorable political environment that offers more activity space for Islamic political parties. In the local elections of 1990, FIS won 31 of the 48 wilayas, and 853 of the 1551 towns. 7 And in December 1991, the party took nearly half the votes at the first round of the legislative election (see Table 2).
For many incumbents in the military and government, the scenario of ''les barbus'' 8 becoming the country's decision makers was unimaginable. To prevent the FIS from ascending to power through the second round of election which was almost for sure (Volpi 2006), a coup d'état was staged by the Department of Intelligence and Security (Département du Renseignement et de la Sécurité, DRS) and launched by the Peoples National Army (Armée Nationale Populaire, ANP) to overturn the Chadli's regime and abolish elections. The following months saw a cleansing campaign initiated by the transitional regime which not only banned the party of FIS, but imprisoned thousands of its militants and sympathizers on concocted charges, denying their access to legal representation. The military's harsh violence against the FIS reinforced large numbers of Islamists faith that Islamic renaissance can only be achieved at the battlefield rather than at the ballot box. The country was then mired into a brutal warfare between the army and radicalized Islamists that lasted for 10 years.

Controlled inclusion and electoral Islamists struggle
The coup of 1992 gave rise to a fierce dispute within the regime itself between the ''eradicators'' and ''conciliators''. The ''eradicators'' referred to the senior military officers that initiated the putsch. Ideologically, this circle intended to eradicate ex-FIS members and the armed Islamist groups, while the ''conciliators'' valued the importance of dialogs with these forces and felt anxious about limitations of the indiscriminate repression strategy (Roberts 2003). Beyond that, a deeper divergence was rooted in the competition for state power. The coup of 1992 brought the military officials to the front of the political scene and the transitional regime institutionalized the army's prominent political role by allowing senior officials' presence in the High Security Council (Haut Comité d'É tat, HSC) and National Transition Council (Conseil National de Transition, CNT), two institutes newly set up to replace the president and the parliament that were dissolved by the military. Even after the transitional period ended in 1995 when the presidential system and plural party election were restored, military elites continued to manipulate the country's policy-making at the backstage and refused to give up their ruling power ( Table 3).
The conciliators first represented by the FLN's Abdelhamid Mehri, then by President Liamine Zeroual and Abdelaziz Bouteflika, always sought to seize back the power and get rid of the military's intermingling in politics.
The fact that Algerian incumbents were themselves part of the field of political competition raises important implications for Islamists non-involved in FIS and violence such as HMS and MRI. Throughout the 1990s, the eradicators sought to use the two moderate Islamic political parties against their competitors within the regime.
To transform the Islamic political parties from opposition forces to their allies, eradicators adopted the ''stick and carrot'' policy. On the one hand, an atmosphere of horror was created at the political scene. By brutally sabotaging FIS's access to the ruling power through elections and cruelly torturing its partisans, the eradicators set a warning to alert other Islamic political parties that they could have the same fate if they refused to make a compromise. On the other hand, access to the country's executive was used as baits to attract Islamic forces. In the immediate aftermath of the coup in 1992, HMS was offered a place in the CNT. Also, Sassi Lamouri, a member of an HMS affiliated organization was nominated minister of religion in 1992 and 1993 (Hamladji 2002). Chin. Polit. Sci. Rev. (2016)  Such a combination of hard and soft methods worked effectively to ally HMS with eradicators. Instead of taking a face-to-face confrontation with the army, Nahnah came up with the moucharaka (participatory) strategy and regarded it as ''a way out of the Algerian crisis'' (Hamladji 2002). Yet, in HMS's eyes, participation did not equal to collaboration or cooperation. As a high-profile party cadre later commented, ''we did not have another choice and we did not want fighting'', 9 they framed the participatory approach as an expediency made in a particular context. The intention was to remain to be part of the political game by making certain compromises while keeping itself in abeyance as an opponent. And the approach was to let the party act as a political influencer by taking cabinet portfolios and pushing the regime to rectify the anti-Islamic policies or behaviors such as corruption.
In the context of an increasing bloody war, HMS's discourse aroused an echo with a large proportion of the ex-FIS's constituency who started to rethink the tragic scenarios in the past few years. The idea of boosting changes via peaceful and inconspicuous means was particularly welcomed by the middle class voters. In the Presidential election of 1995, HMS's head, Nahnah, was second only to the Zeroual. He won 2,971,974 votes (J.O.R.A. (Journal Officiel de la Republique Algerienne) 1995), which accounted for over a quarter of the valid votes. Bearing in mind that the FIS acquired a little more than 3 million votes in the first round of legislative election in 1991, we can see that HMS appeased a principal part of FIS's social base. It is worth noting, however, that the military incumbents only intended to ensure that the Islamic political parties with which they co-opted were no more than cards serving their interests and posing no challenge to their own primacy. The restored electoral systems were no more than tools exploited by the rulers to hide their authoritarian nature. Thus, the incumbents only wished to invite weak candidates to compete with their preferred figure at elections. Not surprisingly, alarmed by the growth of HMS's popularity, the eradicators became more cautious about the Islamists' participation and designed ways to keep them at control.
Even though more cabinet portfolios were offered to HMS in return for its support, the positions were non-weighty ones such as Minister of Tourism and Craftsmen, Minister of Transports, Secretary of State in charge of environment (Hamladji 2002). And a significant shift was to stop appointing HMS partisans to Minister of Religious Affairs since early 1996.
In 1997, new regulations were introduced to interdict the reference of religion in any political parties' names and platforms. The provisions thus forced HMS to change its title from Hamas to Movement of Society for Peace. 10 By removing the Islamic overtones from HMS and MRI, incumbents aimed to impede them from drawing legitimacy from religion. Prohibited from using Islamic reference and expressed their theological thoughts, Islamists found it harder to create an ''Islamic defender'' image and declare the government as an apostate in their discourse.
Also, shortly before the Presidential election in 1999, the constitutional council denied Nahnah's qualification to compete in the upcoming Presidential election because he lacked a proof of ''participation in the Algerian War of Independence''. The disqualification was seen as an intentional action aiming to oust Nahnah, a figure that had the potential to become a real competitor, from the electoral game (Willis 2002: 6). In fact, there were so many similar tricky restrictions to all competitors of the military preferred Bouteflika that all opposition candidates dropped out before the election started, making Bouteflika the only candidate to be elected (Bouandel 2001).
Compared with HMS, MRI was a nut harder for the eradicators to crack. Unlike Nahnah, Djaballah ''steadfastly remained in the opposition'' (Parks 2013: 121) following the coup. Even though he did not involve his party in violence taken by armed Islamists, he criticized the military's intervention in electoral process and call for the re-legalization of FIS (Willis 1998).
Although MRI was not able to mobilize a support base as large as that of HMS in the late-1990s, its unyielding position helped the party gain support from a small proportion of electorate that remained to be sympathizers of FIS. In the parliamentary election of 1997, MRI acquired 8.7 % of the vote share (915,446 out of 10,496,352, see Table 2), and became the fourth largest party at the parliament after RND-a newly founded party of President Zeroual, HMS, and FLN-then controlled by the military which replaced the party's conciliatory secretary-general Mehri with Boualem Benhamouda, an elder military cadre. 10 In Arabic, Movement of Society for Peace is written as Harakat Mujtami'a al-Silm. So the party always called itself HMS. Chin. Polit. Sci. Rev. (2016) 1:717-735 727 As Djaballah's discourses and actions were generally in favor of the conciliators, the military elites took more coercive policies to his party than to HMS. Apart from forcing MRI to follow the rules of ''removing political parties' religious symbols'' as HMS did and delete the term of ''Islamic'' from the party's original name of ''Movement of the Islamic Renaissance'', invisible tactics were taken as a complement to narrow the political openness for MRI. For example, Djaballah and his MRI were intimidated from time to time. Some members even suffered from ''physical harassment'' (Willis 1998).
Meanwhile, internal strife within MRI was staged by incumbents as a camp against the party's charismatic figure, Djaballah, was deliberately created. And in this play of division, Lahbib Adami who had a mysterious background was chosen as the hero. Adami was a high-rank cadre of MRI and was known as a deputy elected from the wilaya of Khenchela in the parliament. Interestingly, he was also brother of the then Minister of Justice, Mohamed Adami (Roberts 2002). In the late 1990s, Adami openly challenged Djaballah by pointing out the negative consequences and the party's losses caused by Djaballah's anti-military strategies. His program was to learn from the experience of HMS and take a more proeradicators' position, which, in his argument, would bring more benefits to the party of MRI. In concert with Adami's attack, an intensification of harassments and threats to MRI's partisans helped waver Djaballah's charismatic position in the party. With the backing from his brother and other elites, Adami successfully evicted Djaballah out of the party in 1999 and swiftly changed the party line from opposition to co-option with eradicators.

Further deterioration in Bouteflika's era
The disqualification of Nahnah's candidacy on the eve of the presidential election in 1999 deprived him of the chance to stand on the electoral arena himself. Nahnah did not have many choices but to work with the new presidential candidate, Bouteflika, who promised to give his party a reward and fulfilled his promise by offering HMS three important cabinet positions after he was elected (Driessen 2012).
Although Bouteflika was himself was a veteran of the ANP and was the presidential candidate designated by the eradicators, he devoted himself to the career of national reconciliation, the idea of which was inconsistent with the eradicators' will. To alleviate the confrontation between the regime and armed Islamists who still launched terrorist attacks from time to time after the civil war, he implemented the Law on Civil Harmony several months after he ascended power and encouraged radical Islamists to disarm themselves by imposing amnesties.
More importantly, Bouteflika claimed that he would not be a ''three-quarter president'' (Holm 2005) as soon as he assumed power in 1999, indicating that the president would spare no efforts to resist the military's manipulation of political affairs. In the following year, a military reshuffle took place, resulting in the replacement of four regional military commanders who had particular influential role following the coup in 1992.
During the first few years of Bouteflika's era, the president seemed to gain the upper hand in his power struggle with the eradicators. This was marked by the announcement of the then chief of staff in the army, General Mohammed Lamari in 2003 that the military did not have a favored candidate for the upcoming election. And in fact, Bouteflika was the first figure acquiesced by the army to be re-elected as president in the country's history (Holm 2005). Another indication of the President's advantage was the resignation of General Lamari in July 2004. As he was one of the most prominent officers that forced Chadli to resign in early 1992 and eradicated the FIS members afterwards, his resignation caused a stir throughout the country.
In the contest between eradicators and new conciliators led by Bouteflika, HMS stood side by side with the latter. In 2004, the party became part of the presidential alliance to support Bouteflika's re-election. In return, Bouteflika added two more cabinet portfolios to the party.
In spite of adopting the same strategies as it did in the late-1990s, HMS experienced a sharp decline in its popularity in the 2000s. In the legislative election of 2002, the party merely acquired 7.1 % of the votes (see Table 2), which was less than half of the vote share it gained in 1997 and almost a quarter of the supporting rate won by the party in the presidential election of 1995. The party's performance in the legislative election of 2007 improved just slightly and merely received an extra of 32, 937 votes (see Table 2).
At least two reasons could be identified to interpret the HMS's fading: Firstly, fierce rifts arose in the party after its founder, Nahnah, died in 2003. The focus of the dispute was who should become the successor of Nahnah. Aboudjerra Soltani who took the leadership after Nahnah's death was born in 1954 and belonged to the second generation of HMS members. It is of notice that ''a number of the second generation leaders had similar profiles''. 11 This means that none of them could distinguish him and earned a charismatic position like Nahnah did. Among the leaders who challenged Soltani's succession was his deputy, Abdelmadjid Menasra, who finally departed the party with sixteen of HMS's deputies and thousands of partisans in 2009 (Parks 2013).
Secondly, as the party got more involved in the cabinet work but failed to bring any real change to the government's policies, its discourse of standing in the ''third way between regime and opposition'' (Willis 2012) became untenable. Corruption hardly reduced after HMS took more cabinet portfolios and parliamentary seats. Also, even though the party claimed itself as a religion defender, its weak performance on the religious issues such as the family code revision made such a declaration questionable. In 2004, a revision of the family code in practice since 1984 was carried out by Bouteflika. The aim was to secularize the code and abandon regulations such as the one that required the presence and approval of the bride's male tutors in a marriage. Despite expressing a rejection voice at the very beginning, HMS soon softened its tone and stated that it would accept the reform of code as long as Bouteflika lifted the state of emergency (MSP: Le marché islamiste 2004). And in the end, HMS actually welcomed the reform even though the state of emergency remained in place.
As for the party of MRI, Adami's putsch against Djaballah finally led the latter to leave the party and create a new party called Movement for National Reform (Movement du Rénouveau National, MRN). Djaballah's departure brought large numbers of his supporters and followers transferred their allegiance to his new party. Meanwhile, the MRI shrank swiftly despite its effort to curry favor with Bouteflika, as the party's new leader, Adami, lacked the prestige and qualifications of Djaballah. In the legislative election of 2002, MRN acquired a supporting rate of 9.5 % and took 43 seats in the parliament, while the MRI received a minimal support and merely gained 1 seat. Yet, the growth of MRN did not last long. Since Djaballah took a hard-line position towards Bouteflika after he became president and openly criticized his policies such as the family code revision, the president adopted the same strategy as the eradicators did and responded to Djaballah by catalyzing divergences within MRN. In March 2007, two months before the legislative election, the scenario of 1999 repeated itself and Djaballah was once again expelled from the leadership by his partisans.
The political landscape transformed dramatically in the late 2000s. The eradicators-conciliators competition turned to the advantage of the former. The change was partly caused by the deterioration of the president's health. Bouteflika was reported to get cancer in 2008 (Arous 2011) and appeared much less frequently in public. This offered the eradicators a chance to return to the front of the political scene. A series of clandestine actions were taken to undermine the conciliatory group. And the struggle culminated in 2010 by the investigation of Sonatrach's corruption initiated by Muhammad Mediène, head of the country's intelligence institute, DRS. Sonatrach was the largest government-owned hydrocarbons enterprise and accounted for 98 % of the country's foreign currency receipts (Migdalovitz 2010). The company had long been controlled by members of Bouteflika's circle in the 2000s. As a result of the investigation, several core entourages of the President including three Vice Presidents and two ministers were dismissed (Del Panta 2014).
By the early 2010, the power wrestling within Algeria's ruling elites basically ended with the triumph of the eradicators, an important sign of which was the remarkable increase of the country's military expenditure (Entelis 2013;Del Panta 2014).
With the downfall of conciliatory elites, a key element of political opportunities for electoral Islamists shifted dramatically: elite divisions. As the eradicators took the hegemony of the regime, they neither had the necessity to use Islamic political parties as counterbalance powers to resist their adversaries within ruling elites, nor had the intention to re-cooperate with the electoral Islamists such as HMS which betrayed them and joined Bouteflika's camp. The transformation from a divided political environment within the ruling elites to an almost undivided one helped the incumbents to reach a consensus on restricting the Islamist electoral force.
A reduction of benefits acquired from the incumbents epitomized by HMS's loss of important cabinet positions reinforced the centrifugal tendencies between elites and Islamic political parties. The Islamist victory in Egypt, Tunisia and Morocco also brought strong demonstration effect for Algerian electoral Islamists. In early January 2012, HMS, MRI and MRN decided to establish an Islamic electoral coalition called AAV as an opponent force against the regime.
Evidently, the original intention of the alliance was to unify the Islamist resources of the country. Nonetheless, given the acute divisions within Algerian electoral Islamists, this goal seemed to be an illusion which could never be achieved.
When Soltani called for the four ministers of HMS to resign their cabinet positions prior to the election to show the party's real breakaway from the presidential alliance, he received negative replies (''Political Context of Algerian Islamist Party's Exit from Presidential Alliance'' 2012). Then, the Minister of Public Works, Amar Ghoul, in particular, was encouraged by the incumbents to openly stand against HMS's withdrawal from the government. Within HMS, Ghoul's idea aroused sympathy from at least 2500 mid and high level members (Boufatah 2013).
As for MRN, the party suffered from more divisions after its founder, Djaballah, was expelled. Lots of its members left the party and joined Djaballah to set up his new party, Justice and Development Party ( On the other hand, rifts among AAV's three component parties made the alliance even more fragile. HMS, the largest party within AAV and initiator of the Islamist alliance, showed contempt for its two partners. And it was not uncommon to find HMS members complaining that ''the other small parties dragged us down''. 12 Apart from the complicated divisions among electoral Islamists, their inability to build a legitimate image in 2012 added to their disadvantage. Alerted by the Arab Spring that started with uprisings in Tunisia, the Algerian military-dominated incumbents unified themselves to stay in power. To prevent the Islamists from taking advantage of mass protests, the ruling elites developed the discourse of ''Islamic parties equal to terrorist groups'' which severely damaged Islamists' mobilization capacity as an electoral force. The civil war in 1992-2002 was generally remembered as a horrifying nightmare for Algerian people. More than 150,000 lives were lost in the decade (Ashour 2009) and people all over the country witnessed murder every day. When the battle reached a climax, it was not uncommon to see an extremist killing his sibling just because the latter served in the police. As time went on, collective trauma of the bloody war developed into an emotion of collective fear and hatred, which could release powerful energy once inspired.
Against the backdrop of the Arab Spring and rise of Islamists in Arab countries, Algerian incumbents used media as loci to disseminate an Islamophobia by repeatedly bringing up the subject of the black decade and by reinforcing the scenario of ''FIS's return'', ''Islamists' occupation'', ''Algerians must be prepared for the rise of religious power''. As an Algerian scholar has mentioned, ''for over a decade since Bouteflika became President, the authority was used to talk about the future in their discourse. But after the outbreak of the Arab Spring, the regime started to talk more about the history of the war in the 1990s. The purpose was to frighten people''. 13 Additionally, they reinforced the concepts that all Islamist tendencies were the same in nature. The Islamists who claimed themselves to be peaceful just wanted to ''hide their game and never say what they think'' (Aziri 2011). The true intention of Islamists is only to expel all those who do not follow sharia and set up a dictatorial rule (Arrêt des É lections de 1991: Objectifs et Effets sur l'Opposition Algérienne 2012). The discourse was such a strong one that none of the Islamic political parties was able to rebuild a legitimate image to a large social base when the election of 2012 was held. Their fiasco then became inevitable.

Conclusions
The decline of Algerian Islamists as an electoral force was largely determined by three reasons: (a) the decrease of political openness caused by the transformation of ruling elites' power landscape from a divided environment to an undivided structure; (b) the fragmentation of Islamic political parties; (c) Islamists' failure to build a legitimate image by their discourse. Hence, arises the question: does this mean that Islamic political parties have no chance at all to rise again in Algeria? Indeed, existing Islamic political parties such as HMS, MRI, MRN, FJD, FAN and so on are so fragmentary that even if favorable opportunities are in place again, they could hardly become the largest beneficiary. Yet, we cannot exclude the possibility of a brand new Islamist electoral force that sets up a better organized party and normative discourses according to the lessons it learned from its predecessors. And it seems that the opportunities for Islamist forces are increasing these days since a new round of power wrestling has started again since Bouteflika's ousting of the DRS's head, Mohamed Mediene, in September 2015.
The case of Algeria suggests that a reference of Islam per se cannot directly bring popularity for a party. Rather, a religious discourse can only be transformed into mobilization strength when tactically used by social movement entrepreneurs to convince the constituency why and how those religious elements would bring a change. Moreover, the preceding research highlights the necessity of focusing on the interplay between political contexts, organizational structures and framing process. Contrary to the descriptive attention based on a socio-economic perspective, the political process model offers a more panoramic picture of Islamic political parties' ebb and flow. Lastly, the case indicates the importance of examining institutionalized Islamist electoral force as a component of social movement theory. To understand Islamic political parties first as social movements then as individual electoral forces has profound policy and theoretical implications.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.