Abstract
Under the Justice and Development Party AKP and Recep Tayyip Erdoğan, Turkey has become one of the most polarized countries in the world, and has undergone a significant democratic breakdown. This article explains how polarization and democratic breakdown happened, arguing that it was based on the built-in, perverse dynamics of an “authoritarian spiral of polarizing-cum-transformative politics.” Furthermore, I identify ten causal mechanisms that have produced pernicious polarization and democratic erosion. Turkey’s transformation since 2002 is an example of the broader phenomenon of democratic erosion under new elites and dominant groups. The causes and consequences of pernicious polarization are analyzed in terms of four subperiods: 2002–2006, 2007, 2008–2013, and 2014–present. In the end, what began as a potentially reformist politics of polarization-cum-transformation morphed into an autocratic-revolutionary one. During this process, polarization and AKP policies; the politicization of formative rifts that had been a divisive undercurrent since nation-state formation; structural transformations; and the opposition’s organizational, programmatic, and personal shortcomings fed and reinforced each other.
Numerous studies—for example, on political party polarization, voter preferences, and social distrust—characterize Turkey as one of the most socially and politically polarized countries in the world (Erdogan 2016; Aydın-Düzgit and Balta 2018; Aytaç, Çarkoğlu, and Yıldırım 2017; KONDA 2017; Yılmaz 2017; Erdogan and Uyan Semerci 2018; McCoy, Rahman, and Somer 2018). Accordingly, the greater part of the political discourse and, in critical elections and referendums in recent years, the lion’s share of voter behavior appears to have been frozen into two mutually disagreeable and obstinate blocs.
For instance, in the 2014 and 2018 presidential elections, where voters and election campaigns were divided almost evenly between those adamantly supporting and fiercely opposing President Recep Tayyip Erdoğan, he won with 51.79 and 52.59 percent of those voting, respectively. 1 Similarly, constitutional changes replacing Turkey’s formal parliamentary system with an executive presidential system passed in a 2017 referendum with 51.41 percent of the voters approving the changes. The remaining portion of the body politic was highly mobilized against the changes, arguing that the new system would formally end democracy and establish strong-man autocracy. The high participation rates in these votes—74.13 percent in 2014, 85.43 percent in 2017, and 86.24 percent in 2018—also suggested an increasingly polarized and politically mobilized electorate. 2
The country is now widely viewed as having undergone a democratic breakdown and degenerating into an autocracy (Özbudun 2014; Diamond 2015; Esen and Gumuscu 2016; Somer 2016; Freedom 2018; Levitsky and Ziblatt 2018). This is striking because Turkey had long been hailed in the world as having built a rare, long-standing, developing, and exemplary case of secular democracy in a Muslim-majority society, while its many serious shortcomings were also noted (Rustow 1970; Özbudun 1996; Hale and Özbudun 2010; Turan 2015). What is more, many scholars had praised Turkey for being on the path of consolidating liberal democracy, and they had commended the Justice and Development Party (AKP) and its leader, Erdoğan, for building a “Muslim (or conservative) democracy” in the twenty-first century (Özbudun 2006; Kuru and Stepan 2012; Yavuz 2009). Hence, a seasoned observer had argued in 2008 (though also citing the warning signs): Turks have at long last begun winning the civic revolution they have been waging for decades. Turkey’s democratic institutions have proven strong enough to contain and guide this revolution, allowing it to proceed peacefully and within the bounds of law. This is a transcendent vindication of the system shaped by revolutionaries of the 1920s, and of the Turkish Republic they created. (Kinzer 2008, xiii–xiv)
What happened, and what can we learn from the Turkish case regarding the relationship between polarization and political regime change? My goal in this article is to answer the question by examining how and why polarization emerged and became pernicious in Turkey, and how and why this contributed to democratic erosion, through several periods during the past two decades. How does polarization help to explain the utter collapse of democratic institutions in recent years, which Kinzer had applauded in 2008 for their ability—despite their inadequacies—to maintain an ongoing peaceful and lawful “civic revolution”? In turn, how did democratic backsliding reinforce polarization? What options have been available for depolarizing the country and for protecting and reforming democracy?
I also discuss a related question of bottom-up versus top-down polarization. To better understand both the Turkish case and the phenomenon of polarization at large, it is crucial to explain to what extent, and in what sense, one can attribute Turkey’s polarization to its “formative rifts” (Somer and McCoy, this volume), in particular a “center-periphery” or “religious-secular” division, which emerged during late-Ottoman and early-republican processes of modernization. After all, Samuel Huntington once ominously described Turkey as a “torn country” (Huntington 1996). He referred to well-known and long-existing social-cultural cleavages (Mardin 1973; Berkes 1998; Kalaycıoğlu 2012), but, by treating them as if they were culturally and historically given and fixed, he also implied that they were decisive.
In fact, the level of politicization and impact on polarization of these cleavages are variable over time. The critical question to explain is how such divisions in some periods become, or, more accurately, how they are made the basis of a “pernicious” type of polarization, i.e., a polity’s division into mutually distrustful “us vs. them” blocs (McCoy, Rahman, and Somer 2018; Somer and McCoy, this volume). For example, even though the fault lines Huntington and others described remained the same, polarization varied across time, as extant measurements of political party polarization also indicate. During the 1980s and 1990s, the Turkish party system had been called “a borderline case between moderate and polarized pluralism” (Ozbudun 1981, 234; Kalaycıoğlu 1994). By contrast, in 2015, Turkish politics was the most polarized among the thirty-eight countries included in the Comparative Study of Electoral Systems (CSES) data, based on Dalton (2008)’s index of party system polarization (Erdogan and Semerci 2018, 37–38).
In response to the above questions, I argue that Turkey’s transformation under the AKP and Erdoğan since 2002 fits with, and helps to further develop, theory regarding a certain causal pattern: “polarization (leading to) democratic erosion or collapse under new elites and dominant groups” (McCoy, Rahman, and Somer 2018; Somer and McCoy 2018).
Further, I maintain that a great deal of democratic backsliding in Turkey can be explained by the built-in perverse dynamics of what I call the “authoritarian spiral of polarizing-cum-transformative politics.” The very same polarizing tactics the AKP employed to mobilize a winning majority, undermine the opposition, and overcome societal and institutional resistance to its transformative policies—and to redistribute power, status, and resources to its own bloc—triggered changes that transformed the AKP itself and the mainstream political field at large. Together with the responses of the opposition and state institutions, this logic of polarization locked both the party and its rivals in a web of intended, unintended, and mutually reinforcing policies and discourses, which were antidemocratic or had democracy-destroying consequences.
In the end, what initially looked like reformist polarization with democratizing potential was transformed into revolutionary polarization, which had destructive, uncertain, and uncontrollable implications for Turkey’s democratic regime. This suggests an important indicator of an authoritarian spiral turning pernicious: when polarization begins to irreversibly transform the polarizing political actor itself.
Transformation through Polarization in Turkey
Echoing the causal pattern of polarization leading to democratic backsliding under new groups (McCoy, Rahman, and Somer 2018; Somer and McCoy 2018), the AKP represented “new” (i.e., from the outskirts of the mainstream) political actors who had previously been marginalized, mainly because of their background as political Islamists. Turkey’s secular laws and pro-secular institutions and elites—the country’s “center”—had treated them with suspicion and some disdain. Though they had been allowed to participate in the political system, they had periodically been reprimanded with legal and political sanctions (Somer 2007 and 2014; Hale and Özbudun 2010).
How have these political outsiders (relatively speaking) come to dominate Turkey’s politics, society, and, increasingly, economy? They did so by coalescing and mobilizing a winning coalition from a diverse, cross-class and cross-ideological base of elite and constituency support (Çarkoğlu and Kalaycıoğlu 2009; Ocakli 2015). A financial meltdown in 2001, among other factors, provided opportunities, and the party managed to use different degrees of polarizing politics in different periods, as I elaborate below. Hence, it mobilized diverse groups based on a simplified and polarizing framing of Turkish society and by promising wide-ranging changes in political and economic structures.
Over time, and with growing polarization, this coalition formed into a partisan bloc that was increasingly personified in Erdoğan. It thus amassed sufficient weight to challenge the well-established institutions of the Turkish “strong state” (Heper 1992; Migdal 2001; Somer 2016).
When the AKP came to power, Turkey’s legacy of a strong state—which had successfully overseen modernization and transition to a multiparty, partial democracy but had also prevented the establishment of full democracy during the twentieth century—had long been in need of reform (Heper 2002; Kinzer 2008; Somer 2016). Strong states can impede democratization (Slater 2012). In Turkey, democratic reforms were necessary, for example, to ensure a transparent and accountable state and to redesign the state-religion relationship based on societal consensus.
Previous elected governments had failed to push through reforms because, among other reasons, they had been too fragile in terms of their support base and vis-à-vis their political rivals, an activist and rowdy media, and the guardian state institutions such as the military and the judiciary. By comparison, the AKP gathered sufficient power and stability to make wide-ranging changes, as I elaborate below, partly thanks to its polarizing-cum-transformative politics.
Many of the earlier changes secured economic growth, made potentially democratizing reforms, and subdued the meddlesome military. The latter was achieved through a mix of legitimate democratic and “new authoritarian,” underhanded, or at times outright criminal, methods (Cizre and Walker 2010; Somer 2016; Yavuz and Balcı 2018). The party could often get away with this because its (over time, perniciously) polarized and captive constituencies in civil society and politics were willing to overlook and sometimes actively support these policies.
During this process, both the AKP and its antagonists often justified their positions in terms of democracy. Similar to other cases in this volume, pro-Erdoğan Turks saw democratization and promise in the same developments in which anti-Erdoğan Turks saw authoritarianism and decay.
At the end of the day, the cost of all this for Turkish democracy and society has been high. In addition to severe polarization, in 2018 Freedom House downgraded Turkish democracy from “partially free” to “unfree” for the first time since the 1980–83 military regime (Erdogan and Semerci 2018, 37–38; Freedom 2018). Accordingly, the three elections and one referendum the AKP has won since November 2015 have not been free and fair, because, among other reasons, plummeting media freedoms and building of a party-state unleveled the playing field against the opposition (Esen and Gümüşçü 2017; Freedom 2018). All in all, democracy broke down under new elites and dominant groups (Özbudun 2014; Esen and Gümüşçü 2015; Taş 2015; Somer 2016; Öktem and Akkoyunlu, eds 2017; McCoy, Rahman, and Somer 2018).
Analyzing how this happened helps to uncover the causal mechanisms through which polarization becomes pernicious and undermines democracy. I identify ten mechanisms, which I will summarize in the last section, by tracing how four causally important periods unfolded:
- 2002–2006: Trimodal, moderate, and “micro-textual” polarization, and reforms led by the AKP. Significant interparty cooperation.
- 2007: Micro-textual polarization culminates in a political confrontation. Beginning of “macro-textual” polarization.
- 2008–2013: Growing macro-textual polarization, decreasing trimodal and reversible polarization. Incremental democratic erosion. Opposition polarization emphasizes secularism. AKP side highlights the “new Turkey and reforms” theme as an upper text, but Islamism exists as a subtext. Trimodal polarization ends with the 2013 political confrontation.
- 2014–present: Full-fledged bimodal and pernicious (self-propagating) polarization spiraling out of control and democratic backsliding. Polarization is increasingly personalized and ossified into pro- and anti-Erdoğan camps. Islamism-secularism frame continues as a subtext.
But polarization under AKP governments was not created from a tabula rasa. The party capitalized on what might be called Turkey’s “protean formative rifts,” which took on different names and forms when represented by different actors in different periods.
Turkey’s Two Formative Rifts
Periodic polarization and polarizing politics have been part of Turkish politics since late Ottoman times, based on two types of formative rifts. The first rift concerned “outsiders” in Turkey’s historical processes of modernization and “Turkish” nation-building, i.e., the cultural, religious, and linguistic groups that the ruling elites saw (or who saw themselves) as too different to be part of their sociocultural and political-economic projects. Hence, minority groups such as non-Muslims, Kurds, and Alevis were either denied citizenship or treated as citizens with suspicious loyalties unless they culturally assimilated into state-defined “Turkishness” in public life. Hence, the state and ruling elites used polarizing politics to mobilize the Turkish-speaking Sunni Muslim majority population and to justify policies against these other groups (Göçek 2011; Öktem 2011). 3 This type of formative rift might have produced a historical legacy, mode, and repertoire of polarizing politics that current governments can tap into.
However, my main concern—because of its more direct link with the current polarization—is a second type of rift that researchers have described with terms such as “center-periphery,” “state-society,” and “secular-religious” (Mardin 1973, 2006; Bozdoğan and Kasaba 1997; Berkes 1998; Findley 2010). This rift is a division within the Turkish-speaking Muslim majority, i.e., the “insiders” and main “target population,” so to say, of the state-led processes of modernization and nation-building.
Until the AKP period, one side of this rift consisted of Turkey’s transformative, developmental, and authoritative central state and the institutions and social-political groups that were its primary defenders and beneficiaries. On the other side were societal segments that felt overlooked and objectified by the state and by the people whom they perceived as enjoying a closer and more favored relation with the state. For lack of better terms, let us call these two sides “republican-centrist” and “provincial-revisionist,” respectively.
Even though the “haves” of society were by definition overrepresented in the first and the “have-nots” were overrepresented in the second—education and formal employment being the key determinants of upward social mobility and socialization into state-favored lifestyles and identities—both categories have cut across class boundaries. Initially, urban-dwellers and the agents of state-led modernization such as civil servants and teachers were the mainstays of the first category (Mardin 1973). But both groups became more diverse and dynamic through socioeconomic modernization, rural-urban migration, and more market-based growth, after the 1980s (Bozdoğan and Kasaba 1997). This rift also cuts across ethnic-linguistic divisions, except perhaps for unassimilated ethnic Kurds (Demiralp 2012).
At first sight, the rift seems to overlap with a Left-Right division, because right-wing political actors have predominantly represented provincial-revisionist groups. Yet, periodically, leftist politics (including pro-Kurdish parties) has also represented it with some success, for example during the 1970s and early 1990s. Finally, the cultural-ideological nature of the “center” has changed depending on which political groupings control the state (Somer 2014a; Bilgin 2018).
This legacy produced two competing “foundational” narratives of Turkish modernization. The centrist right-wing and left-wing versions of these narratives have not been mutually exclusive. Both upheld the main myths and truth-claims of Turkish nation-building during the 1920s and 1930s, the difference being that right-wing actors aimed to curb its excesses through more religion- and market-friendly policies (Somer 2014a). However, these narratives also had versions, held by Islamist and secularist elites, that more or less excluded each other (Somer 2010).
Furthermore, before the AKP, power was to some degree balanced between the political representatives of the two sides. In party politics, right-wing political parties dominated the system. The Republican People’s Party, CHP—originally founded by Atatürk—and other “left-wing” parties could rule only in coalition governments with right-wing parties. But the secular-republican ideology was dominant in the military, the judiciary, and various other state institutions, as well as the mainstream media, civil society, and big business. These horizontal and vertical forces balanced the power of right-wing elected governments.
These discursive ideological (partially overlapping narratives) and institutional (power balance) factors help to explain why pre-AKP polarization did not become pernicious despite the presence of formative rifts. For example, the Left-Right ideological polarization of the 1970s did not produce two ossified blocs as we see today; on the contrary, volatility and fragmentation were recurring features of the post-1960s party system (Sayarı, Musil, and Demirkol 2018).
The AKP was rooted in political Islamism, whose foundational narratives were in many respects mutually exclusive with the republican-centrist narratives, and they represented “new actors” with a “passive-revolutionary” and “state-conquering” agenda (Tuğal 2009; Somer 2010 and 2017). Yet, initially, the AKP formed a center-right coalition based on a more inclusive upper-text or “macro-text.” This macro-text bundled center-right narratives with themes such as globalism, economic development, Muslim Turkish nationalism, neo-Ottomanism, anti-elitism, and EU membership (Öniş 2007; Çarkoğlu 2008; White 2014), while Islamist narratives continued as a subtext or “micro-text” (Somer 2007).
Causes, Dynamics, and Periods of Polarization under the AKP
Polarization is a multidimensional and relational process, and we need indicators that measure it as such (Lauka, McCoy, and Firat 2018). The causal narrative below focuses on qualitative signs and processes. Simultaneously, it is worthwhile to note that extant measurements of party-ideological polarization also indicate that Turkey became increasingly polarized during these periods. 4
2002–2006: Moderate and micro-textual polarization and reforms
The AKP came to power in 2002 when major segments of the Turkish electorate had grown weary of the existing political elites and were ready to support large-scale changes, following a decade of political-economic instability. The first AKP period in government generated only moderate and what I call “micro-textual” polarization, because polarization in this period could be described as trimodal or bisected. In the middle of those who approved and disapproved of the party, mainly because of its Islamist credentials, stood many elite and societal groups. These groups lent their conditional backing to the new elites because they saw them as agents for change, EU membership, and ending military tutelage. Thus, these groups and many external observers were willing to overlook subtle but deeply polarizing micro-text discourse and behavior from the AKP (Tepe 2005; Somer 2007, 2010; Kinzer 2008).
Unlike the inclusive, reformist, compromising, and thus nonpolarizing upper-text, this polarizing subtext was “passive-revolutionary” (Tuğal 2009). It was revealed by behaviors and expressions at the local level and oblique statements at the national level (Somer 2007), and it reflected an orientation toward gradual state-capturing and redefining society based on Islamist foundational narratives (Somer 2010, 2017). These factors raised ontological insecurity among pro-secular groups (Somer 2007; Öktem and Akkoyunlu 2017).
Reinforcing these undercurrents were “state-conquering” government policies—e.g., gradual staffing of state agencies with partisans, including members of the Gülen Islamist movement (Gülen Hareketi [GH]), and subtle signs of Islamization in social life (Somer 2007; Toprak et al. 2008; Öktem and Akkoyunlu, eds 2017; Yavuz and Balcı 2018). On the macro-level, the government’s discourse and policies were shaped by reformism, democratization, and EU membership (Özbudun 2007; Hale and Özbudun 2010). Many legal-political reforms were passed through inter-party cooperation. Reflecting the trimodal polarization, although pro-Islamic and pro-secular elite views in the press diverged on issues such as secularism, religion’s role in society and state, and social pluralism including gender rights, they converged on issues such as political democracy—neither group being “exemplary democrats of a principled and inclusive kind”—and each group exhibited significant internal diversity (Somer 2010b; Somer 2011, 514).
2007: Political confrontation and the beginning of “macro-textual” polarization
At this critical juncture, pernicious polarization and authoritarianism could have been prevented in the decade ahead if Turkey’s “old” political elites could have united and reformed the discourses, programs, and campaign strategies of the opposition parties, thus democratically checking and balancing the AKP’s growing power (Somer 2007; Kumbaracıbaşı 2009). But they performed these tasks rather poorly. Instead, they resorted to mass polarizing politics and legal measures, which often had questionable constitutional legitimacy, to remove or discipline the AKP in a context of economic growth and AKP popularity.
Furthermore, secularists drew on the Islamism vs. secularism cleavage at a time when this division was growing in importance among significant pro-secular segments of society but might have been subsiding among a majority consisting of religious conservatives and moderate secularists. Reflecting the latter trend, in a 2006 survey, about half the respondents—presumably mostly AKP supporters—thought the party “believes in democracy (53.7 percent) … defends fundamental rights and freedoms more than other parties (50.8 percent) [but protects] an Islamic way of life more than other parties (53.3 percent).” Moreover, the proportion of those believing that religious people were repressed had fallen from 42.4 percent in 1999 to 17.0 percent in 2006. But other major segments of society thought that the party intended to “impose an Islamic way of life (50.3 percent) . . . seeks EU membership to legitimise [sic] an Islamic political system (45.2 percent) . . . did not soften the Islamist-secularist conflict (44.4 percent) . . . seeks to infiltrate the bureaucracy with Islamist cadres (43.8 percent) and . . . intends to reverse the advances concerning women’s rights (36.7 percent)” (Çarkoğlu and Toprak 2006; Hale and Özbudun 2010, 38–39). There was also a split among pro-secular elites, among whom secularist sensitivities were either increasing or decreasing (Somer 2010b, 568–71).
Then the parliament’s impending election of a new president prompted a political confrontation (Hale and Özbudun 2010, 39). The outgoing president was a secularist and former head of the Constitutional Court (CC) who had obstructed various AKP policies and appointments. Hence, secularists saw him “as a symbol and ‘the last citadel’ of the secular republic [and] reacted strongly to the election of a formerly Islamist politician” (Hale and Özbudun 2010, 40).
In essence, the constitution encouraged “forbearance,” inter-party consensus, and a nonpartisan president by requiring a two-thirds majority of the parliament’s full membership in the first two rounds of the election. Technically, however, it allowed the AKP to elect its own candidate—initially expected to be Erdoğan—because the vote of a simple majority was sufficient in the third and fourth rounds, with no quorum requirement specified for any round.
When the AKP nominated its number two figure, Abdullah Gül, secularist elites mobilized pro-secular members of the middle class in massive antigovernment “republican rallies” (Somer 2007). Further, the opposition CHP went to the CC claiming that a two-thirds quorum must be required during the first round so that the majority party would feel compelled to compromise. Meanwhile, the military issued an online ultimatum accusing the government of undermining secularism and threatening an intervention, amid pro-secular media criticism of the government.
Soon after, the Court upheld the CHP’s case (Hale and Özbudun 2010, 39–40). With these moves, and the AKP’s responses, micro-textual polarization became macro-textual. The AKP mobilized its base against this “affront to democracy” and called a referendum on various constitutional amendments that, among other changes, would explicitly abolish any quorum requirement for the impending presidential election and bring direct popular elections of future presidents. The CC agreed that the proposed changes were constitutional. The AKP carried the vote with 68.95 percent, elected Gül president, and won early parliamentary elections held “under the shadow of the constitutional crisis” with 46.58 percent support (Hale and Özbudun 2010, 40–41).
2008–2013: “Macro-textual” and increasingly bimodal polarization and incremental democratic erosion
This period began with a fierce war of nerves, words, and legal-political maneuvers at the elite level and ended with massive bottom-up uprisings against the government, known as the Gezi Protests, in 2013. Yet, arguably, the authoritarian spiral of polarizing-cum-transformative politics was still reversible, if the opposition parties could reform themselves to increase their electoral weight, link with the growing grass-roots opposition, and employ polarizing politics based on a constructive rather than an obstructive platform.
Early on, two judicial interventions, one from the secularist establishment and the other from the growing Islamist elements within the state, further polarized Turkish society. First was the chief public prosecutor’s case to close the AKP in March 2008. The CC found that the party was guilty of being a “focal point of antisecular activity” but not seriously enough for closure. Instead, the Court issued a financial penalty and a “serious warning,” while simultaneously citing concerns for democracy in the event of banning an elected governing party. The AKP declared this a “victory for democracy” (Tait 2008). However, it became apparent that it was celebrating a triumph for vertical accountability in democracy alone. From then on, and by increasingly framing the necessity of judicial reform as a struggle against “oligarchic rule,” the party implemented policies that incrementally weakened judicial independence and undermined horizontal accountability.
Second, in July 2008, the growing AKP-GH camp within the judiciary opened a series of lawsuits against secularist military officers, intellectuals, and civil society actors, on what later turned out to be mainly trumped-up charges (Cizre and Walker 2010; Jenkins 2011). 5 Coming on the heels of smaller campaigns in 2004 and 2005 against military and secularist actors, including a university rector, and lasting until 2010, these legal maneuvers effectively subdued the secularist armed forces and parts of the pro-secular establishment. Reflecting the continuing trimodal polarization, pro-secular elites in political parties, media, and civil society became split between cynical democrats who viewed these cases—and in general cooperation with the AKP—as necessary to end military praetorianism and alarmists who saw them as anti-secular conspiracies.
The opposition parties could not form a positive and unequivocally prodemocratic agenda uniting the cynics and alarmists. The main opposition party, CHP, strongly condemned the antisecularist trials while sending ambiguous signals about, if not passively approving, the anti-AKP case. The party’s leader supported the legal basis of the AKP-closure case but expressed concerns about the consequences for democracy if the AKP were banned.
Further, opposition parties took an uncompromising and arguably polarizing stand vis-à-vis various legal-constitutional changes that the AKP attempted or passed in this period, arguing that these changes were stepping stones for regime change. Opposition parties had legitimate reasons for concern, among them questions regarding the constitutionality of many amendments. In the case of a 2013 inter-party commission for partially rewriting the constitution, they also maintained that an apparent consensus collapsed in the eleventh hour because AKP insisted on introducing a presidential system, which the opposition believed would be authoritarian.
But the opposition parties appear to have had two main shortcomings. First, they focused on alarmist-obstructive rather than proactive and constructive politics of polarization, thereby neglecting to develop alternative programs of reforms. Second, they did not make any path-breaking changes in their organizations, programs, and electoral strategies, for example, by advancing new methods of civic engagement and political communication, even though the elections were mainly free and fair in this period. Preoccupied with the question of how to stop the AKP’s transformations and inversion of the system’s formal and informal rules, the opposition was trapped in a pro-status-quo position.
By comparison, the AKP managed to maintain an image as the main party that promised change and prosperity, and effectively communicating with its constituencies. The party and Erdoğan employed numerous discursive, coercive, financial, and legal (and at times extra-legal or illegal) methods of polarizing-cum-transformative politics that mobilized their base and weakened their opponents. A comprehensive account is beyond the scope of this article, but one consequential confrontation between the AKP and the media will serve as a representative example.
Early in 2008, Turkey’s then–most influential media corporation, the Doğan Media Group (DMG), began covering news of corruption linked to the AKP, based on convictions in Germany (Daloğlu 2009; Higgings 2009). While the DMG was mainly pro-secular, it also harbored many outlets that were examples of the moderate or “cynical democratic” position. 6
In response, the then–Prime Minister Erdoğan launched a fierce campaign against the DMG, displaying archetypal tools of polarizing-cum-transformative politics. He accused the DMG of spreading false news and blackmailing his government for financial favors, and asked people to boycott it. At the height of his campaign, Erdoğan described the dispute in terms of a broader political conflict taking place in the country, using a textbook example of a “polarizing speech act,” which he later wielded against other critics such as Turkey’s biggest and pro-secular business association, Tüsiad. “Those who do not take sides [in this battle]” he threatened, “will be sidelined.” Only months later, the DMG was charged a record penalty of 500 million dollars for alleged tax evasion (Daloğlu 2009; Higgings 2009).
Acting as a polarizing political entrepreneur and displaying a “rival image” of the contenders in this conflict (Somer 2001; McCoy, Rahman, and Somer 2018), Erdoğan was urging people to interpret political events, and to position themselves politically, in terms of a major rivalry in society. While he invited people to support his own side in the struggle, which according to him would be the winning side, he also strongly discouraged anybody from seeking a middle ground, warning that they would also be targeted and eventually find themselves to be weak and politically irrelevant. As such, his rhetoric had a chilling and Machiavellian undercurrent of politics understood as pure power, where “the power and capacity to determine a friend and an enemy overlaps with the legitimate authority to establish a new legal order” à la Carl Schmitt (Kutay 2018, 1).
Against this background, many moderates and would-be bridge-makers became divided. Some argued that these forceful, confrontational tactics were necessary to reshuffle rigid institutions and dethrone established political-economic elites. Others opined that the government’s means—whatever its ends might be—were dangerous and destructive.
The AKP framed the vote on a second crucial constitutional referendum, which it introduced in 2010, as a battle between those defending their privileges in the “old Turkey” and those supporting a “new Turkey,” similarly polarizing society in a trimodal fashion. The proposed amendments looked like an omnibus bill. They included generally undisputed changes expanding some rights and liberties, such as collective bargaining for civil servants (Kalaycıoğlu 2012). The disputable changes seemed to democratize the election of the high judiciary, but simultaneously cracked open the door to speeding up the progovernment transformation of the judiciary and the military (Kalaycıoğlu 2012, 6; Arato 2010; Cizre and Walker 2010; Jenkins 2011). All of this split the opposition between those supporting the changes as “incomplete but good” and those warning of the end of secular democracy and judicial independence (Kalaycıoğlu 2012).
The AKP government won the referendum with a 58 percent majority, tipping the balance of power in the state from pro-secular to proreligious AKP and GH actors. This outcome was solidified by an electoral victory in 2011.
The period came to a close in 2013 with two polarizing episodes that ended trimodal polarization. The first involved the antigovernment Gezi protests, which were triggered by police violence against young environmentalists in Istanbul. They soon spread throughout the country, gathering millions of supporters. Even though pro-secular sensitivities were a main unifying theme, the participants were a cross-generational and cross-ideological lot brought together by opposition to AKP authoritarianism and the demand for real democracy (Özbudun 2014; Yörük and Yüksel 2014). This grass-roots mobilization offered a golden opportunity for the opposition parties to unify their fragmented constituencies based on an anti-AKP, prochange and prodemocratic politics of polarization. But the major opposition parties—except, perhaps, for the pro-Kurdish minority party—missed this opportunity, arguably because of their ideological divisions and rigidity and their organizational shortcomings as mainly “cartel parties” (Sayarı, Musil and Demirkol 2018). 7 As the AKP tightened its grip on civil society and the media, Gezi was unable to develop into a lasting social movement with a concrete agenda. Hence, it enhanced polarization and solidified the pro-AKP bloc without necessarily strengthening the anti-AKP bloc.
In the second episode, Islamist infighting between the AKP and GH produced a full-blown conflict. Antigovernment forces in the judiciary and police, led by GH-linked elements, brought massive corruption charges against the government (Gümüşçü 2016). The allegations were spread on the Internet through sound and video recordings of party members, including Erdoğan and his family. In addition to increasing Erdoğan’s authoritarianism and accelerating his purge of AKP moderates, this conflict forced more people to choose between two evils: government corruption, on one hand, and antigovernment and illegal formations within the state, on the other. The opposition again failed to open a third, prodemocratic path, encumbered as it was not only by its own deficiencies but also by its disadvantages on an increasingly unlevel playing field.
2014–present: Pernicious polarization and democratic breakdown
While referring the reader to other contributions for the numerous and momentous socio-political developments in this period (see Başkan 2015; Taş 2015; Esen and Gumuscu 2016; Gumuscu 2016; Somer 2016; Esen and Gümüşçü 2017; Öktem and Akkoyunlu 2017; Yavuz and Balcı 2018), suffice it to say here that this period has witnessed incremental backsliding (Bermeo 2016) and collapse of democracy and the emergence of an electoral-authoritarian regime. The AKP’s initial rise as a dominant party gradually gave way to Erdoğan’s personalized rule, with an executive presidential system and a largely instrumental AKP.
Major developments have included: Erdoğan’s election to the presidency in 2014 and his declaration of a “de facto presidential system,” defying the constitutional order based on parliamentarism and a neutral (nonpartisan) presidency; the repeated legislative elections of 2015, where the AKP first lost the majority and then seized it back in unfair and only partially free snap elections; Erdoğan’s de facto efforts to revamp and dominate governmental institutions; the resignation of the elected AKP prime minister under pressure from Erdoğan in the spring of 2016; a failed coup attempt, allegedly led by the GH, in the summer of 2016, after which the government declared a state of emergency and started an anti-GH and anti-opposition witch hunt; a 2017 constitutional referendum establishing a de jure executive presidential system; and snap legislative and presidential unfair and unfree elections in 2018, won by the AKP and Erdoğan despite a highly mobilized opposition.
All these transformations took place in a context of polarizing-cum-transformative politics where the ensuing social-political polarization became increasingly self-propagating, personalized, and based on negative partisanship and fear.
For example, in the 2017 referendum, voters’ preferences vis-à-vis presidentialism were highly partisan (Aytaç, Çarkoğlu, and Yıldırım 2017). With few people understanding the difference between the presidential and parliamentary systems, most voters took their cues from their party preferences, but people with more knowledge of presidentialism were more likely to vote “No.” Religiosity and Kurdish ethnic-linguistic background were significantly and positively correlated with a “Yes” vote, while “No” votes increased along with one’s level of education. These three factors were discussed earlier in relation to Turkey’s formative rift. However, partisanship was not the sole determinant of the vote; satisfaction with the economy was another significant factor in increasing “Yes” votes (Aytaç, Çarkoğlu, and Yıldırım 2017, 9).
What is more, support for presidentialism began to rise primarily “as a result of rising ontological security concerns” after the June election, when a spree of terror attacks broke out and petrified the public (Aytaç, Çarkoğlu, and Yıldırım 2017, 17). Simultaneously, the polarizing-cum-transformative arguments increasingly adopted the form of “either a presidential system or instability and terror.” The “positive” polarizing frame of “either the old elites or new elites and a more democratic Turkey” that had prevailed earlier in the AKP was long gone. It was replaced by a negative frame of “either the new elites or insecurity.”
In this context of full-fledged, pernicious bipolarization, the AKP increasingly saw itself as a hegemonic actor situated in a revolutionary moment and having a mandate to unilaterally rebuild both the state and civil society. Accordingly, AKP-engineered structural changes in the media and economy began more and more to sustain and feed polarization. For example, a decade after the confrontation with Erdoğan discussed above, the DMG—with its owner facing charges—was bought by a progovernment mogul, using a state bank loan with highly favorable conditions. This completed a process begun in 2008, as a result of which most of the media came under the control of progovernment outlets and cronies and the Turkish media became “unfree” (Economist 2018; Freedom 2018; Yeşil 2016). At these outlets, common ground disappeared not only through social, psychological, or intellectual mechanisms but also because journalists and commentators who refused to toe the party’s or leader’s line were sacked or simply no longer given a platform.
As Erdoğan had proclaimed a decade earlier, those who did not want to take sides were indeed left behind. Moreover, the bar of “taking sides” rose as polarizing-cum-transformative politics shifted power to new actors. In the process, many of the earlier agents or apologists of polarization—such as the GH, AKP politicians who refused to become Erdoğan loyalists, liberal writers, and at first independent and then semi-independent journalists and public intellectuals—were marginalized.
As of 2018, Turkey is both politically and socially polarized based on negative partisanship, i.e., how distant voters feel emotionally from the parties they oppose (Erdogan and Uyan Semerci 2018, 38–42), and “perceived threat, but not empathy” shapes “social distance toward Kurds, AKP supporters and AKP opponents” (Bilali, Iqbal, and Çelik 2018, 74). Polarization has transformed the media, public discourse, and social relations. To give just one example, in a 2016 survey, 74 percent of respondents opposed the idea of their children playing “with the children of someone voting for another party” (Erdogan 2016, 2; KONDA 2017).
Conclusions and Ten Causal Mechanisms
Certainly, multiple factors, including ideology and the international environment, contributed to the AKP’s and Turkey’s authoritarian transformation (see Öniş 2012; Grigoriadis 2014; Somer 2017). Notwithstanding those influences, I have argued that polarizing-cum-transformative politics and the dynamics of the resulting pernicious polarization were a crucial part of Turkey’s recent, unfortunate, political history.
The Turkish case helps to uncover the causal mechanisms of pernicious polarization, not only because it tallies with the aforementioned causal pattern but also because it can be analyzed as a theory-disconfirming “crucial case” (Somer 2014b, 3, 7–9). Democratization failed under the AKP despite many auspicious conditions identified in extant theories of democratization (and related theories such as those involving religious actors and democratization). For example, Turkey defies modernization theory, which would not predict that a country with “rising GDP per capita, a burgeoning civil society, and a rising middle class in the past decades … [would] slide into competitive authoritarianism” (Sarfati 2017, 395). It is hard to explain the outcome without taking into account the causal mechanisms of pernicious polarization.
My analysis suggests ten causal mechanisms that were at work when polarizing-cum-transformative politics spiraled out of control, became “normalized,” and replaced what we understand as democratic competition.
First, polarizing politics and discourse empowered opportunistic and revanchist actors within the AKP while weakening more coolheaded actors with a stronger commitment to democracy, pluralism, the rule of law, and compromise. Hence, political actors who could support de-polarization and genuine democratization lost leverage within the AKP bloc.
Second, polarization supported post-truth politics, undermining any existing common ground for democracy. The AKP’s rhetoric vilified the opposition, “old elites,” and the existing political system, often by exaggerating and distorting the truth, if not by fabricating outright lies. These discursive developments crowded out views within the pro-AKP partisan discourse that were relatively more grounded in truth, while justifying more and more authoritarian policies. The polarizing discourse seems to have reached a critical mass (Somer 2001) and gained a self-propagating momentum, perhaps even spiraling out of the control of party leaders—at least of moderate leaders who would have resisted the pernicious kind of polarization.
Hence, third, all these factors increasingly nurtured a captive and partisan constituency. This constituency became increasingly willing to support, overlook, and, at times, demand not only the reformation but also the capturing and revolutionary dismantling of existing democratic institutions and divisions of powers. Captive audiences became increasingly ready to condone growing corruption, violations of the rule of law, and opportunistic grabbing of power and wealth by the party and its clients. 8
This third mechanism triggered the fourth one, which can be called the “dwindling channels of bridge-making.” Relatively nonpartisan agencies and institutions that harbored actors with mixed or noncommitted orientations and thus had the potential to contribute to bridge-building and de-polarization—such as those in the media, business, and civil society—were wiped out. Their ownership and control shifted to progovernment partisan actors as pro-opposition institutions were radicalized. Hence, polarizing-cum-transformative politics became more effective and forceful alongside the creation of crony capitalism, a partisan welfare state, and progovernment civil society and media, which became the financiers, justification, and mouthpieces of such politics (Yoruk 2012; Aytaç 2014; Kaya 2014; Gürakar 2016; Yeşil 2016).
Fifth, as the AKP increasingly captured the state, the vilification of state institutions became a self-defeating strategy. To continue the politics of polarization and maintain their support base, pro-AKP actors had to shift the blame onto other internal and external targets, which included many potential agents of de-polarization.
Sixth, all these policies—and the reactions of the state institutions—meant that progovernment actors increasingly began to perceive the institutions of horizontal accountability, such as an independent judiciary, as a barrier to their goals. This suggests an extension of the argument by Slater (2013) and Slater and Arugay (2018), who see divided societal preferences— favoring either horizontal (elite-institutional) or vertical (i.e., popular) accountability—as a cause of polarization. Rather than, or in addition to, being a source of polarization, aversion to horizontal accountability can develop as a product of polarization. Indeed, while all right-wing Turkish political actors had prioritized popular support over institutional accountability, the more the AKP relied on polarizing-cum-transformative politics, the more it grew hostile toward any autonomous institution enforcing horizontal accountability (Özbudun 2014).
Seventh, polarization encouraged the personalization of politics and the demand for an executive presidential system within the AKP. By definition, polarization simplifies politics to a choice between “either-or” options (McCoy, Rahman, and Somer 2018). With growing polarization, these choices increasingly resemble path-dependent and identity-based attachments, as in fanhood and tribalism, rather than interest- or value-based choices, as in ideological loyalty and preference for a political program.
By its very nature, I would argue, presidentialism—and executive presidentialism—is likely to emerge as the institutional form that represents and organizes a severely polarized polity, because it has a similar logic. Presidentialism and the personalization of politics, too, simplify politics in the form of either-or choices between the personas of a few strong leaders. Hence, it seems to be no coincidence that, as political choices in Turkey became increasingly simplified to a choice of supporting one bloc or the other, they increasingly took the form of either trusting or distrusting President Erdoğan. In this way, the AKP as a party and other contenders for power within the party were overshadowed by Erdoğan’s personalized power (Lancaster 2014). Hence, there may be a causal relation between the synchronous rises of polarization and presidentialism—or rather “one strong man rules” (Stepan 2009; Svolik 2014; Diamond 2015).
Eighth, these developments equally affected the opposition, which found itself between a rock and a hard place. It has oscillated between two strategies since the political confrontation in 2007. Trying to follow a strategy of opposing polarization and seeking compromise—for example, by toning down the anti-Islamist, or, later, anti-Erdoğan rhetoric—often served to legitimize AKP authoritarianism and underhanded political tactics that defied democratic norms. But trying to develop its own polarizing-cum-transformative politics to bring down the government often ended up hardening the AKP camp even further.
For example, when attempting to prevent the executive presidential system the AKP had proposed, whenever the opposition focused on the system’s institutional flaws and authoritarian loopholes, they persuaded their own voters but failed to dent the progovernment voters’ personalized trust in Erdoğan. When they tried to chip away at the trust between Erdoğan and his supporters by targeting his personality and alleged corruption, they again ended up strengthening the progovernment bloc, which closed ranks to protect their leader, and they subjected themselves to the criticism that they were focusing on personality and not proposing alternative policies.
Ninth, similar to cases such as Thailand and the Philippines (Slater and Arugay 2018), the political and bureaucratic elites contributed to the development of pernicious polarization when they tried to remove or discipline the AKP by using strategies that stretched constitutional boundaries—even when these were aimed at punishing the AKP’s own transgressions. The opposition also missed opportunities to reform democratic institutions based on cooperation with the AKP at critical junctures.
Tenth, forced into a pro-status-quo position to obstruct the AKP’s polarizing-cum-transformative politics, opposition parties failed to reinvent themselves as prochange actors with new discourses, programs, and organizational forms (Somer 2007 and 2014b; Kumbaracıbaşı 2009; Ayan Musil 2014). With the opposition now including those who were non-Islamist center-right, doing so could have reestablished the electoral balance of power with the AKP, compelling the latter to act more democratically (LeBas 2011; Somer 2017).
In the end, these mechanisms locked the AKP, its rivals, and potential bridge-makers in a downward spiral of authoritarian politics and democratic erosion. The more polarizing politics proved effective, the more it took control of politics, required more of itself, and weakened democracy.
Until 2014 or so, pernicious polarization was likely preventable. One historical-institutional, one political-economic and organizational, and one agentic facilitating factor seem to have undermined this possibility. Polarization built on the mutually exclusive Islamist and secularist narratives of the formative rift (Somer 2010a). These could be compared with ethnic and national “foundational myths,” which enhance pernicious consequences when they become the basis of polarization (LeBas and Munemo, this volume). Polarizing-cum-transformative politics generated powerful stakeholders in the media, business, and “GONGO society,” i.e., NGOs sponsored or at times directly organized by the government or the governing party, that had political-economic interests in the continuation of polarized politics. Opposition actors performed poorly in uniting and renewing themselves, due to limitations of power politics and ideology, personal shortcomings of self-restraint and talent, insufficient comprehension of the dynamics of polarizing-cum-transformative politics and, thus, an inability to develop novel strategies to control and redirect these dynamics for the purposes of democratic revival.
Footnotes
Notes
Murat Somer is a professor of political science and international relations at Koç University Istanbul. He specializes in comparative politics and democratization, and his research on polarization, religious and secular politics, and ethnic conflict has appeared in books, book volumes, and journals, such as Comparative Political Studies, American Behavioral Scientist, and Democratization.