Erdoğan’s March 31 Elections: A Fiasco of Tactics and Rhetorics

On March 31, Turkish citizens cast their votes to elect local administrators. Municipal elections in Turkey, especially in the big cities, have usually been considered a litmus test for party performance at the national level. This time was no exception: pundits and politicians alike have deemed the results as a major blow to the ruling Justice and Development Party (AKP). According to the unofficial results, the AKP lost the biggest cities to the opposition Republican People’s Party (CHP). However, the narrative is complicated by the fact that the vote-shares of political parties in city councils — better measures of voters’ political preferences in local elections — were much the same compared to the previous municipal elections in 2014.

As we will discuss below, the “People’s Alliance” (PA) of the AKP with the Nationalist Movement Party (MHP) did not suffer a dramatic numerical setback. But nor did the ‘National Alliance’ (NA) of the CHP and the İYİ Parti (İP) achieve a significant increase in its popularity. The March 31 elections instead proved that the electoral success or failure of political actors (parties and alliances) is determined by mastering tactical deployment and mobilizing their more or less stable base support. This is an outcome of the redefinition of the political arena to a battlefield between party alliances in the aftermath of the June 7, 2015 parliamentary elections.

Erdoğan’s failure to foresee the limits of this new front line and his discursive mistake in turning the local elections into a de facto referendum on his regime paved the way for the March 31 fiasco. It is striking how the same combination of alliance policy and campaign rhetoric produced an almost identical outcome to the April 16, 2017 presidential referendum, thereby guaranteeing the most recent setback for the PA. The results, on the other hand, have rendered the AKP more vulnerable to pressures from its allies and to those from within the party. Beyond the AKP’s obvious defeat in this complicated chess game, the MHP emerged triumphant within the losing PA.

The most detrimental actor sealing President Erdoğan’s monumental losses was the alliance of the Kurdish movement and various Turkish leftist groups, represented in the electoral arena by the People’s Democracy Party (HDP). One should particularly add to this picture the most damning defeat of the PA — namely, the loss of Istanbul — which owes a lot to the NA candidate, Ekrem İmamoğlu. An atypically soft-spoken and pragmatic leader, İmamoğlu managed to mobilize support from all sectors, especially the HDP constituency. The ruling elites’ post-election reaction to the shock of losing the biggest metropolitan municipalities, including Istanbul, Ankara, Adana, and Mersin, has been instructive. The foot-dragging in acknowledging the opposition’s victory and manufacturing a pretext to delegitimize the election result in Istanbul (in particular) seems to be triggering a meltdown in mass support for the AKP.


The emergence and deepening of the AKP–MHP alliance


At the June 7, 2015 elections, for the first time in over a decade, the AKP failed to gain a majority after the HDP crossed the 10 percent national election threshold and took 80 seats in the parliament. As a response, Erdoğan opted for a de facto alliance with MHP under the leadership of Devlet Bahçeli, in order to forestall the formation of an anti-AKP minority government by the opposition. The weeks that followed the June 2015 elections witnessed an upsurge of violent conflict between the Turkish army and the Kurdistan Worker’s Party (PKK) militia in the Kurdish-majority east and southeast. Snap elections were held in November 2015, ending in the formation of a single-party government under AKP rule, thanks to votes that the AKP gained from the MHP. Bahçeli has remained a prominent supporter of Erdoğan ever since, backing that was especially visible during the coup attempt of 2016, the constitutional referendum in 2017, and Turkey’s military operations in northern Syria between 2015 and 2018.

For Erdoğan, the pragmatic motivation of the coalition with Bahçeli has been to remain in power; for Bahçeli, it has been to secure state support against his rivals within the MHP (who later abandoned the party to launch the İP) and to re-infiltrate the state bureaucracy. Ideologically, the alliance has been fused by the deep opposition the two share to collective Kurdish demands in Turkey — but also in Syria and Iraq. Turkey’s miscalculated foreign policy in Syria and the concurrently changing power configurations on the ground — have provoked severe territorial anxieties among a range of political actors in Turkey, but most notably the ultra-nationalists. These anxieties were catalyzed when the Kurdish Democratic Union Party (PYD) and its armed wing, the People’s Protection Units (YPG), became key players in the US-led struggle against ISIS in 2014. The abortive coup attempt in July 2016 further exacerbated these anxieties, turning them into an existential crisis over the survival of the Turkish state.

The de facto alliance between the MHP and the AKP became formal heading into the 2018 presidential and parliamentary elections. Erdoğan’s win in the presidential elections was made possible by the passing of favorable (to him) election bill on March 14, 2018. Yet, the real surprise of the 2018 elections was the MHP’s performance. Beating almost all the pollsters’ forecasts, MHP — a party that did not even hold election rallies — gained 11.1 percent of the vote, preserving its vote share from the November 2015 snap elections. In a speech he delivered after the elections, Devlet Bahçeli noted that the “Turkish nation has not only brought the party to a key position within the parliament, it also gave the party a major responsibility to balance power.”

Yet, an unintended consequence of Erdoğan’s alliance with Bahçeli has been to drive the scattered opposition together, producing a broad-based (if reluctant) electoral bloc. This bloc showed its electoral potential for the first time in the presidential referendum of 2017, which Erdoğan won with a bare margin. Unbowed, the NA recruited a new partner — the Islamist Felicity Party (SP) —to try to strengthen the opposition bloc for the June 24, 2018 elections. This second attempt also failed to topple Erdoğan but it also indicated the limits of his popular support and his dependence on the MHP. Erdoğan won the presidency, yet, because of Turkey’s election system (blocs+threshold), he fell short of securing an absolute majority in the parliament, consolidating Bahçeli’s role as kingmaker in Turkish politics.


Erdoğan’s dependency on Bahçeli increases after March 31


After an initial phase of ambiguity and uncertainty following the 2018 general elections, the AKP and MHP eventually agreed to continue their ballot alliance in the 2019 local elections. Bahçeli stated on August 31, 2018 that the MHP did not want to lose its gains in the upcoming March elections, and that “a return to the pre-July 15, 2016 conditions should be prevented by dissolving through the will of the people the evil alliance between CHP and HDP and other substitutes.” Throughout the campaign period, pro-regime actors worked in concert to cast the elections in apocalyptic terms. According to this rhetoric, the survival of the regime, the country’s territorial integrity, and even the nation’s very existence were at stake. This discursive tactic thus turned regular municipal elections into a de facto referendum on the legitimacy of the ruling alliance nationally.

In hindsight, this was a major mistake. A close look at the distribution of municipalities by political parties (and thus party alliances) shows a trend in electoral preferences that parallels the results of 2017 referendum, in which AKP deployed a similar tactical alliance and campaign rhetoric.  Coastal cities and Kurdish cities that voted ‘NO’ in the referendum voted against the AKP–MHP alliance in the 2019 local elections as well.

Moreover, the continuation of the alliance and the discursive tactics that it employed, seemed to benefit the “junior” MHP far more than the AKP as the “senior” partner. The MHP won 11 municipalities, up from the eight municipalities that it won in the previous municipal elections in 2014. Moreover, of these eleven municipalities, seven were taken from the AKP. In addition, in 30 cities in Anatolia, where the AKP and the MHP did not run together but competed against one another, the AKP’s vote share in city councils decreased and that of the MHP increased.

Bahçeli has also arguably further increased his bargaining power within the PA after March 31. In a speech he delivered on April 3 about the Istanbul recount, the NA candidate, İmamoğlu, called on Bahçeli, together with Erdoğan, to do the right thing. İmamoğlu noted that he was expecting Bahçeli, based on his long experience as a statesman, to intervene to prevent intra-group struggles driven by personal concerns, ambitions, and failures from undermining the elections. Blunting İmamoğlu’s call, however, Bahçeli made a public show of alliance unity the next day, asking for necessary measures to be taken against “those people and networks who prevented recounting of the votes and enabled fraud.” Since then, there has been a systematic propaganda campaign organized by various pro-government newspapers and think tanks such as SETA circulating the same narrative, calling Istanbul elections an “organized crime”, and even a “coup.” On April 7, the co-chair of the AKP also joined this chorus in a press speech where he announced that the party would seek an official recount of all the votes in Istanbul due to suspicions of “organized malfeasance.”


The shock over Istanbul exposes intra-party frictions


The March 31 elections seem to have exposed two co-centric and interrelated spheres of tension in the PA: the increased grip of the MHP and Bahçeli on the regime, and; deepening intra-AKP struggles. For sure, given the atmosphere of fear within the bureaucracy, it is impossible to know the exact nature and extent of these tensions. Yet, we can still draw some tentative inferences, especially in light of the increasingly chaotic tension over the winner of the Istanbul metropolitan municipality.

The contestation over Istanbul started in the early hours of Sunday evening when the state’s new agency, Anadolu Agency (AA), stopped sharing election data. At that time, only 24 percent of the ballot boxes had been counted, with Binali Yıldırım — the People’s Alliance’s candidate — registering 51.83 percent of the vote. Around the same time, Yıldırım tweeted, from his personal account, a photo taken while preparing his “thank you” speech with a note that he had won. In the following two hours, İmamoğlu appeared multiple times in front of the cameras accusing the AA of manipulating the results, since less than one-quarter of the ballot boxes had been counted. AA resumed sharing data shortly before midnight. At the time when Yıldırım had declared victory earlier in the evening, only 40 percent and 70 percent of ballot boxes had been counted in Beşiktaş and Kadıköy, respectively, two key pro-CHP districts (and among the most populous) in Istanbul. To everyone’s surprise, Erdoğan appeared shortly after to give a speech in which he appeared to concede, noting that “even if the people delivered a metropolitan municipality overall [to the opposition], they gave the districts to the AKP.”

The chaos of the night was further intensified when the pro-government daily newspaper Sabah reported that 30 polling clerks had been taken into custody in Istanbul based on fraud allegations. Other pro-government outlets also shared the news on their websites. Yet, the Istanbul Governorate, the Mayor’s office, and the Police Headquarters all refuted this. To add even further to the confusion, Istanbul streets were covered in the early hours of April 1 with Yıldırım’s posters thanking the people for his supposed “victory”. Around 11 am, the Supreme Electoral Council (YSK) announced that İmamoğlu was leading the race in Istanbul. AA confirmed the YSK data two hours later. At the time of writing, invalid ballots in Istanbul are still being re-counted.

Interestingly, even though the dominant narrative in pro-government outlets about the Istanbul elections has turned into one of systematic fraud — i.e., a threat to the country — over the course of a week, few columnists in the pro-government papers, Yeni Şafak and Yeni Akit, raised criticism of the party.


The HDP’s game-changing election strategy


There is no doubt that had it not been for the support of HDP, İmamoğlu would not have succeeded in gaining more votes than Yıldırım in Istanbul. In fact, this is also the case for all other metropolis, perhaps except Izmir, where the CHP–İP candidate beat the AKP–MHP candidate handily. Comparing the vote share of parties in city councils with the vote share of alliances in metropolitan municipalities makes it clear that the HDP played a crucial role. Even though the NA parties made it explicit that the HDP was not welcome in the alliance, the HDP elected to openly support NA candidates in those districts where it chose not to run itself. This decision was, without doubt, a game changer in Turkey’s new politics of ballot alliance. The straw that broke the camel’s back was the call — one week before the elections — by the HDP’s jailed leader, Selahattin Demirtaş, for HDP supporters to go to poll stations and vote for anti-PA candidates regardless of their political discourse/position.

Reflecting on the HDP’s election performance in Kurdish cities and towns, the tactical war of alliances makes it extremely difficult to come up with precise and valid comparisons and inferences. The party’s performance in the heartland varied across electoral districts. The HDP was able to reclaim some of the municipalities in which the party’s elected governors had been replaced by government-appointed trustees (kayyum) after the 2016 coup attempt. In the three highly-populated metropolitan municipalities of the Kurdish region (Diyarbakır, Van, Mardin) the party increased its vote share (by 7.2, 0.3 and 2.4 percentage points, respectively). Leaving aside the provinces in which the HDP supported other party candidates, the party’s support increased in eight provinces and decreased in six. Yet, the most dramatic decrease came in the two provinces that have been the strongholds of the Kurdish movement for the past two decades, Şırnak and Hakkari. Nevertheless, the party did retain supremacy (although much reduced) over the combatant AKP in both these provinces, although the AKP took the mayoralty of Şırnak from the HDP.

One should bear in mind that the HDP’s performance in these municipal elections should be evaluated with the regional political opportunity structure in mind. Şırnak and Hakkari have suffered by far the most from the urban destruction that has racked the region since the June 7, 2015 elections and have also been subject to drastic demographic shifts. These are also the provinces in which state repression and attempts to liquidate Kurdish solidarity networks (including anti-poverty NGOs) has been the most fervent. In other words, these are the provinces in which locals are most in need of the financial support resources that the central government controls.

In short, Şırnak and Hakkari were model provinces for the AKP government and its interior minister, Suleyman Soylu — a hardliner in the government, who ambitiously and repeatedly vowed to uproot the PKK and its supporters once and for all. In the end, the PA (especially the AKP) appears to have won a great pyrrhic victory in “conquering” the Şırnak municipality. For it has done so without showing any significant gains in the metropolitan Kurdish municipalities and has paid a hefty price in losing Istanbul, Ankara, Adana, Mersin, and Antalya in its campaign against the HDP.


A paradigmatic shift in the significance of the ballot in Turkey


The ongoing frenzy within the ruling PA over the recounting of votes has left many observers inside and outside of Turkey in doubt about the continuing salience of the ballot box in Turkey. Most pundits seem to forget, however, that nullification of the voting tallies was already foreshadowed during the election campaign. The three leading figures of the PA — Erdoğan, Bahçeli and Soylu —all repeatedly and incessantly promised that they would scrutinize the poll results before “deciding” who would be allowed to take up municipal posts. The increasing distance between voting results and governing “outcomes” began in the wake of the June 7, 2015 elections — a critical juncture in the path towards institutional disintegration and authoritarian consolidation of power. This path was sealed at the elections of June 24, 2018. As we reflect on the March 31 elections, we will likely remember them as a second critical juncture making the path difficult to reverse.

The difficulty lies not only in the increasing arbitrariness of exercising power, bad enough in and of itself. More importantly, it is driven by a paradigmatic shift in what casting a ballot is supposed to signify in Turkey. To say the least, voting is increasingly seen — at least by the ruling clique — as “indicative” and not “determinative”. The undemocratic move to appoint government trustees in many Kurdish cities after 2016 reflected this discourse, as it was rhetorically based on perceived “patriotism” and “loyalty” to Turkey — the elected Kurdish mayors were charged with supporting or fomenting terrorist activities and organizations. This discourse of “indicative” ballots by the three PA leaders was clear for all to see in the run-up to March 31, when they declared that only “the right kind” of person, irrespective of the vote tally, should fill mayoral offices across the country. This indicated the post-2016 logic would now apply beyond the Kurdish parts of the country. The remarks made by the PA leaders throughout the campaign about the NA candidate for Ankara mayor, Mansur Yavaş, are clear indications of this.

The post-March 31 crisis has, for the first time in Turkish history, put the sufficiency of ballot results as the foremost source of legitimacy for representation in Turkey in question. But this was, as we have intimated above, foreshadowed by the nature of post-election conflicts leading back to the June 7, 2015 poll. What was not expected, apparently — especially by the PA —was İmamoğlu’s victory in Istanbul. Since Erdoğan had not anticipated the win, he had not thought to set İmamoğlu up in advance of the poll and so made no effort to paint him as not “the right kind of person” to hold office. In any case, as of March 31, Turkey has entered a new era — one that will be perilously uncertain but that, we predict, will offer new opportunities for the pluralist and progressive forces in the country to rise again.

Photo: Ozan Kose / AFP

Follow us on Twitter, like our page on Facebook. And share our contents.

If you liked our analysis, stories, videos, dossier, sign up for our newsletter (twice a month). 



Please consider giving a tax-free donation to Reset this year

Any amount will help show your support for our activities

In Europe and elsewhere
(Reset DOC)

In the US
(Reset Dialogues)