Reading the Black Sea Tea Leaves: Post-Referendum Analysis

170411160412-02-cnn-turkey-referendum-0411-yes-istanbul-referendum-karafox-04-full-169

Speaking before supporters, celebrating yet another electoral victory, President Erdoğan sounded many familiar themes: he promised to serve the nation rather than be its master; he emphasized the historic nature of the referendum and the heroism of those who had voted in support of his priorities; and he ended by leading the crowd in a recitation of his favorite song, “We Walked Together on Those Roads.”[1] There were also darker, newer additions to his stump speech: he made sure to thank minor (albeit extreme) nationalist and religious parties that had supported him; he dismissed attempts by opponents (nearly 49% of votes) to “diminish” the electoral accomplishment; and he acknowledged the crowd’s bloodthirsty cries for a reintroduction of the death penalty by promising to hold a referendum if necessary.[2]

These appeals to nationalism and the worst inclinations among voters are precisely what worry so many of Erdoğan’s critics. Not only is the country in an official state of emergency marked by sweeping purges and vicious terrorist attacks, but the president is also using the sort of violent rhetoric one expects to hear from would-be authoritarians in countries like the Philippines or United States. It is this tendency to dismiss or demonize opponents in conjunction with the new powers the referendum gives Erdoğan that have led so many commentators in America and Europe to prophesy the “end” of democracy in Turkey. Others point out the that electoral campaign has not been “free or fair,” which suggests that the victory was largely preordained, merely a reflection of a “slide” into dictatorship already occurring.[3]

Yet, even if the Erdoğan’s victory was a foregone conclusion, the electoral results reveal a great deal about the political situation in Turkey.[4] The binary choice of the referendum obscures a number of challenges now facing President Erdoğan and his political allies.

HOW TO UNDERSTAND THE VICTORY OF THE YES CAMPAIGN.

Ostensibly, the YES campaign is supported by an array of parties of which the Justice and Development Party (AKP) is merely one. The most significant of these parties is the Nationalist Action Party (MHP), but while the leadership of the MHP supported the referendum, the rank-and-file may not have. As for the other parties, they have miniscule followings in the country. The best comparison, then, is between the AKP’s most recent electoral performance in November 2015 and Erdoğan’s presidential victory in 2014.

Nationwide, Erdoğan won the 2014 presidential campaign with 51.8% of the vote—just enough to avoid a run-off. That election was a three-man race in which his opponents only had regional appeal. As the only candidate with national appeal Erdoğan eked out a majority. Contrasting 2014 with 2017 helps understand the extent to which voters who were willing to elect Erdoğan president were willing to give him greater powers.

Similarly, comparing AKP vote share in November 2015 with the YES vote in the referendum is instructive. After experiencing a decline in support in June 2015, the party worked assiduously to demonize the Peoples’ Democracy Party (HDP) and peel off votes from the Nationalist Action Party (MHP). As a result, in the November election the AKP received 49.49% nationwide (only slightly off the 49.95% it received in 2011). By comparing this result to the YES vote, we can get some sense of how many voters, after voting for AKP, were willing to grant further powers to Erdoğan.

In both cases, one might expect some drop-off and, sure enough, both comparisons show precisely such an outcome—and something more too: the largest declines from the presidential vote in 2014 to the referendum vote in 2017 occurred along the Black Sea littoral—a region where Erdoğan and his allies have consistently done well. To be sure, the referendum still passed by large margins, but the decline was large as well. In Sinop, for example, the referendum passed with 57.8%—but that was 3.4% less than what Erdoğan received in 2014. And, although 73,324 voted in favor, that is 425 less voters than supported the AKP in 2015 (when turnout was a percentage point higher), and 857 less than in 2014 (when turnout was considerably lower).

Put differently: since taking office, and despite terrorism and coup attempts, President Erdoğan has been unable to mobilize some of his most vocal supporters to grant him greater powers.

As for the provinces where the YES vote was higher than the November 2015 AKP vote or the 2014 Erdoğan vote, these occurred almost entirely in the southeast. While it is possible that voters in this region have grown sick and tired of PKK militants using urban areas as a base and provoking government reprisals, the numbers suggest an alternative: namely, that two years of violence and the government’s crackdown on regional political organizations have made mobilization difficult.

Moreover, with around 400,000 citizens in the region displaced, reaching the ballot box is nigh impossible.[5] In Şırnak, where PKK-government clashes have been intense, over 10,000 fewer citizens voted than in November 2015. While the province rejected the referendum by 71.7%, the YES percentage was significantly higher than either the AKP (12.3%) or Erdoğan (14.8%) had previously received.[6]

Significant Gains 2014 to 2017 (Larger than 3%)
Erdoğan (2014) Referendum: YES % Change
Hakkâri 16.4 32.4 16
Muş 35.6 50.6 15
Şırnak 14.8 28.3 13.5
Osmaniye 48.6 57.8 9.2
Kars 42.5 51 8.5
Yozgat 65.8 74.3 8.5
Iğdır 26.9 34.8 7.9
Bingöl 65 72.6 7.6
Bitlis 52 59.4 7.4
Ağrı 36.4 43.1 6.7
Erzurum 68.8 74.5 5.7
Siirt 42.4 47.8 5.4
Tunceli 14.4 19.6 5.2
Mardin 36.6 41 4.4
Ardahan 40.7 44.3 3.6

Erdogan Gains

Significant Declines 2014 to 2017 (Larger than 2%)
Erdoğan (2014) Referendum: YES % Change
Artvin 52.6 46.9 -5.7
Ordu 67 61.9 -5.1
Giresun 66.6 61.7 -4.9
Karabük 64.6 60.1 -4.5
Bolu 66.1 62.3 -3.8
Trabzon 70.1 66.5 -3.6
Zonguldak 53 49.4 -3.6
Sinop 61.2 57.8 -3.4
Uşak 50.1 47 -3.1
Edirne 32.2 29.5 -2.7
Çanakkale 41.9 39.5 -2.4
Ankara 51.3 48.9 -2.4
Karaman 66.2 63.9 -2.3
Balıkesir 47.8 45.5 -2.3
Samsun 65.9 63.6 -2.3
Izmir 33.4 31.2 -2.2

Erdogan Loses

Significant Gains 2015 to 2017 (Greater than 3%)
AKP (11/1/15) Referendum: YES Change
Hakkâri 12.6 32.4 19.8
Şırnak 11.1 28.3 17.2
Muş 33.9 50.6 16.7
Ağrı 26.9 43.1 16.2
Bitlis 43.8 59.4 15.6
Kars 35.6 51 15.4
Van 30 42.7 12.7
Mardin 28.5 41 12.5
Siirt 37.7 47.8 11.1
Yozgat 64.8 74.3 9.5
Batman 28.4 36.4 8
Bingöl 64.7 72.6 7.9
Tunceli 11.7 19.6 7.9
Gümüşhane 68.3 75.2 6.9
Erzurum 68.1 74.5 6.4
Şanlıurfa 64.6 70.8 6.2
Ankara* 43.5 (1)/55.5 (2) 48.9 5.4 (1)/-6.6 (1)
Adana 36.8 41.8 5
Mersin 31 36 5
Çankırı 69.1 73.5 4.4
Iğdır 30.9 34.8 3.9
Erzincan 56.8 60.5 3.7
Tokat 59.6 63.2 3.6
Ardahan 36.1 44.3 3.6
Çorum 61.3 64.5 3.2
Isparta 53.2 56.4 3.2

* Ankara is divided into two voting districts for parliamentary elections, see below for breakdown of Ankara by municipality.

AKP Gains

Most Significant Declines 2015 to 2017 (Greater than 0.5%)
AKP (11/1/15) Referendum: YES Change
Ankara* 43.5 (1)/55.5 (2) 48.9 5.4 (1)/-6.6 (1)
Istanbul* 47.5/51.2/47.8 48.6 1.1/-2.6/0.8
Kilis 65.6 64.1 -1.5
Konya 74.4 72.9 -1.5
Ordu 63.2 61.9 -1.3
Denizli 45.5 44.5 -1

* Ankara is divided into two voting districts for parliamentary elections; Istanbul into three. See below for breakdown of Ankara and Istanbul by municipality.

AKP Declines

 

Some of the most significant gains came in Europe where the bulk of Turkey’s 2,929,389-stong diaspora lives. Voting from abroad has only been possible since 2013 and the low turnout among diasporic citizens in 2014 suggests that politicians had not yet worked to mobilize them. In Holland, for example, where 8.6% of diasporic voters were registered at the time of the referendum, turnout was only 7.2% in 2014, but rose to 46.7% and 46.8% in November 1, 2015 and April 16, 2017, respectively. Though Erdoğan’s share of the vote in 2014 (78%) was higher than in either the AKP (69.7%) or the YES vote (70.9) received, the difference in turnouts makes comparison difficult. Moreover, Holland was part of a larger trend in 2017: of European countries with more than 10,000 diasporic voters only voters in Switzerland registered a decline in support for Erdoğan’s priorities.

 

 

AKP Gain in Foreign Countries With +10,000 Voters (Nov. 2015-2017)
  As % of Foreign Voters AKP% 11/1/15 (Turnout%) Referendum: YES% (Turnout) Change%
Denmark 1.2 50.1 (41.4) 60.6 (33.3) 10.5
France 11.1 58.4 (45) 64.9 (43.8) 6.5
Sweden 1.3 40.7 (34) 47.1 (26.6) 6.4
Belgium 4.7 69.4 (40.1) 75 (53.1) 5.6
Austria 3.7 69 (40.6) 73.2 (48.6) 4.2
Germany 48.8 59.7 (40.8 63.1 (46.2) 3.4
Holland 8.6 69.7 (46.7) 70.9 (46.8) 1.2
Italy 0.5 36.8 (40.1) 37.9 (40) 1.1
United Kingdom 3.2 20.2 (38.6) 20.3 (38.7) 0.1

 

 

 

AKP Decline in Foreign Countries With +10,000 Voters (Nov. 2015-2017)  
  As % of Foreign Voters AKP% 11/1/15 (Turnout%) Referendum: YES% (Turnout) Change%
Switzerland 3.3 45.7 (9.9) 38.1 (44.8) -7.6
Canada 0.9 34.6 (33.1) 27.9 (46.1) -6.7
Australia 1.5 47.8 (24.1) 14.8 (32.4) -6
USA 3.4 20 (26.7) 16.2 (32.9) -3.8
Saudi Arabia 0.8 58.7 (22.6) 55.1 (32.4) -3.6
North Cyprus 3.6 49.3 (34.2) 45.2 (41.6) -4.1
Greece 0.3 25.2 (6.1) 22.6 (7.6) -2.6
Russia 0.4 27.2 (25.9) 26 (28.7) -1.2

 

 

Erdoğan Gains in Foreign Countries With +10,000 Voters (2014-17)
  As % of Foreign Voters Erdoğan% 2014 (Turnout%) Referendum: YES (Turnout%) Change%
Belgium 4.7 69.9 (6.3) 75 (53.1) 5.1
USA 3.4 15.9 (11.8) 16.2 (32.9) 0.3

 

 

Erdoğan Declines in Foreign Countries With +10,000 Voters (2014-17)
  As % of Foreign Voters Erdoğan% 2014 (Turnout%) Referendum: YES (Turnout%) Change%
Saudi Arabia 0.8 80.6 (6.5) 55.1 (32.4) -25.5
Greece 0.3 44 (4) 22.6 (7.6) -21.4
Australia 1.5 56.4 (10.5) 14.8 (32.4) -14.6
North Cyprus 3.6 54.9 (12.1) 45.2 (41.6) -9.7
Italy 0.5 45.8 (7.8) 37.9 (40) -7.9
Holland 8.6 78 (7.2) 70.9 (46.8) -7.1
Austria 3.7 80.2 (10) 73.2 (48.6) -7
Canada 0.9 33.4 (8.7) 27.9 (46.1) -5.5
Germany 48.8 68.6 (8.1) 63.1 (46.2) -5.5
Sweden 1.3 51.1 (4.2) 47.1 (26.6) -4
United Kingdom 3.2 23.5 (6.4) 20.3 (38.7) -3.2
Russia 0.4 28.7 (10.3) 26 (28.7) -2.7
Denmark 1.2 62.7 (3.4) 60.6 (33.3) -2.1
Switzerland 3.3 39.6 (9.9) 38.1 (44.8) -1.5
France 11.1 66 (8.4) 64.9 (43.8) -1.1

 

 

 

 

 

 

WHAT IS THE SIGNIFICANCE OF YES LOSING IN ISTANBUL AND ANKARA?

A common—and not wholly wrong—observation following the referendum was that the NO victory in Turkey’s three largest cities was a rebuke to Erdoğan. Perhaps, but it requires some caveats. In the first place, Erdoğan and the AKP only won a plurality of votes in the 2014 presidential election and the November 2015 election. In Istanbul in 2014, Erdoğan won 49.8% of the vote with a turnout of 72.8%.[7] In 2015, the AKP won 48.72% with a turnout of 88.1%.[8]

In 2017, the YES vote garnered 48.65% of the Istanbul vote with a turnout of 88.7%.[9] In other words, even in these urban areas where NO won, the change was small. In effect, a fractional number of voters who were willing to vote for the AKP over other parties and Erdoğan over other candidates were unwilling to give him sweeping powers.

The loss in the cities, therefore, should not come as a surprise. What is suggests is an inability of the Erdoğan and his government to expand their appeal. Beyond that, it suggests Erdoğan cannot even bring all of his traditional supporters along with him: comparing Istanbul municipal results we see that in the municipalities containing 62% of the Istanbul population, the referendum received a lower share of the vote than the AKP won in November 2015; in the municipalities containing 75% of Istanbul’s population, the YES vote share was lower than Erdoğan’s vote share in 2014.[10] In Ankara, the numbers are more eye-opening: in municipalities containing 44.2% of the population, the YES number was lower than the November 2015 AKP vote share; in municipalities containing 99.9% of the population, the YES was lower than the Erdoğan’s 2014 vote share. [11][12][13]

Istanbul Results by Municipality (2015 vs. 2017)
% of Population[14] AKP (11/1/15) YES Change
Sultanbeyli 2.2 66.9 70.5 3.6
Arnavutköy 3.3 63.4 66.7 3.3
Bağcılar 5.1 59.6 61.1 1.5
Esenyurt 5.4 46.1 47.5 1.4
Sancaktepe 2.5 50.4 51.7 1.3
Başakşehir 2.5 52.8 53.9 1.1
Esenler 3.1 65.4 66.4 1
Sultangazi 3.5 60.6 61.4 0.8
Avcılar 2.9 40.4 41 0.6
Şile 0.2 50.9 51.4 0.5
Silivri 1.2 39.9 40.1 0.2
Çatalca 0.5 39.7 39.8 0.1
Küçükçekmece 5.2 44.4 44.4 0
Bahçelievler 4.0 50.4 50.3 -0.1
Çekmeköy 1.6 52.6 52.4 -0.2
Ümraniye 4.7 55.5 55.2 -0.3
Gaziosmanpaşa 3.4 57.4 57 -0.4
Zeytinburnu 1.9 51.3 50.9 -0.4
Güngören 2.0 54.3 53.8 -0.5
Kartal 3.1 44 43.5 -0.5
Büyükçekmece 1.6 44.8 44.3 -0.5
Tuzla 1.6 51.3 50.7 -0.6
Sarıyer 2.3 41.6 40.9 -0.7
Fatih 2.8 52.2 51.4 -0.8
Kağıthane 3.0 54.9 54.1 -0.8
Eyüp 2.6 49.3 48.5 -0.8
Beyoğlu 1.6 50.9 50.1 -0.8
Beylikdüzü 2.0 41.9 41.1 -0.8
Adalar 0.1 27.2 26.3 -0.9
Ataşehir 2.9 43.6 42.6 -1
Şişli 1.8 29.3 28.2 -1.1
Beşiktaş 1.3 18.2 17.1 -1.1
Maltepe 3.3 39.5 38.4 -1.1
Üsküdar 3.6 47.9 46.7 -1.2
Beykoz 1.7 54.6 53.1 -1.5
Kadıköy 3.1 21 19.4 -1.6
Bakırköy 1.5 23.9 22.3 -1.6
Bayrampaşa 1.8 55.4 53.1 -2.3
Istanbul Results by Municipality (2014 vs. 2017)
% of Population 2014 Erdoğan YES Change
Arnavutköy 3.3 65.1 66.7 1.6
Sultanbeyli 2.2 69.6 70.5 0.9
Bağcılar 5.1 60.6 61.1 0.5
Başakşehir 2.5 53.5 53.9 0.4
Esenyurt 5.4 47.1 47.5 0.4
Sancaktepe 2.5 51.6 51.7 0.1
Sultangazi 3.5 61.4 61.4 0
Esenler 3.1 66.5 66.4 -0.1
Silivri 1.2 40.4 40.1 -0.3
Avcilar 2.9 41.4 41 -0.4
Küçükçekmece 5.2 45 44.4 -0.6
Çatalca 0.5 40.5 39.8 -0.7
Bakırköy 1.5 23 22.3 -0.7
Beşiktaş 1.3 17.8 17.1 -0.7
Çekmeköy 1.6 53.3 52.4 -0.9
Umraniye 4.7 56.3 55.2 -1.1
Bahçelievler 4.0 51.4 50.3 -1.1
Büyükçekmece 1.6 45.6 44.3 -1.3
Beylikdüzü 2.0 42.4 41.1 -1.3
Sariyer 2.3 42.3 40.9 -1.4
Pendik 4.7 58.7 57.2 -1.5
Zeytinburnu 1.9 52.4 50.9 -1.5
Güngören 2.0 55.4 53.8 -1.6
Kartal 3.1 45.1 43.5 -1.6
Şile 0,2 53.2 51.4 -1.8
Eyüp 2.6 50.4 48.5 -1.9
Gaziosmanpaşa 3.4 59 57 -2
Şişli 1.8 30.3 28.2 -2.1
Kadıköy 3.1 21.6 19.4 -2.2
Ataşehir 2.9 44.8 42.6 -2.2
Üsküdar 3.6 49.1 46.7 -2.4
Kağıthane 3.0 56.5 54.1 -2.4
Adalar 0.1 28.7 26.3 -2.4
Fatih 2.8 54 51.4 -2.6
Maltepe 3.3 41.4 38.4 -3
Beyoğlu 1.6 53.2 50.1 -3.1
Beykoz 1,7 56.2 53.1 -3.1
Bayrampaşa 1.8 56.3 53.1 -3.2
Tuzla 1.6 54.7 50.7 -4
Ankara Results by Municipality (2015 vs. 2017)
% of 2016 Population 2015 AKP YES Change
Evren 0.1 58.5 63.8 5.3
Çamlıdere 0.1 80.6 84.4 3.8
Bala 0.4 64.1 67.2 3.1
Kızılcahamam 0.5 67 69.7 2.7
Akyurt 0.6 76.8 79.4 2.6
Şereflikoçhisar 0.6 54.9 57.5 2.6
Haymana 0.5 69.5 71.6 2.1
Kalecik 0.2 64.3 66.2 1.9
Nallıhan 0.5 52.4 54.3 1.9
Kahramankazan 0.9 68.2 69.8 1.6
Çubuk 1.6 75.7 77 1.3
Elmadağ 0.8 45.6 46.8 1.2
Güdül 0.2 67.1 68.2 1.1
Pursaklar 2.6 75.5 76.4 0.9
Beypazarı 0.9 59.9 60.7 0.8
Sincan 9.7 64.8 65.3 0.5
Mamak 11.7 49.5 49.9 0.4
Keçiören 17.0 55.5 55.6 0.1
Altındağ 6.8 63.2 63.3 0.1
Gölbaşı 2.3 52.8 52.7 -0.1
Etimesgut 10.1 45.6 45.5 -0.1
Yenimahalle 12.1 42.7 42.5 -0.2
Ayaş 0.2 54.8 54.4 -0.4
Çankaya 17.2 22.4 21.7 -0.7
Polatlı 2.3 56.5 54.3 -2.2
Ankara Results by Municipality (2014 vs. 2017)
% of 2016 Population 2014 Erdoğan YES Change
Çamlıdere 0.1 83 84.4 1.4
Akyurt 0.6 79.4 79.4 0
Haymana 0.5 71.7 71.6 -0.1
Nallıhan 0.5 54.5 54.3 -0.2
Kahramankazan 0.9 70.6 69.8 -0.8
Pursaklar 2.6 77.8 76.4 -1.4
Bala 0.4 68.7 67.2 -1.5
Sincan 9.7 66.8 65.3 -1.5
Çubuk 1.6 78.6 77 -1.6
Şereflikoçhisar 0.6 59.2 57.5 -1.7
Yenimahalle 12.1 44.4 42.5 -1.9
Mamak 11.7 51.9 49.9 -2
Etimesgut 10.1 47.6 45.5 -2.1
Elmadağ 0.8 48.9 46.8 -2.1
Keçiören 17.0 57.7 55.6 -2.1
Altındağ 6.8 65.5 63.3 -2.2
Çankaya 17.2 24 21.7 -2.3
Güdül 0.2 70.6 68.2 -2.4
Evren 0.1 66.5 63.8 -2.7
Kızılcahamam 0.5 72.4 69.7 -2.7
Beypazarı 0.9 63.6 60.7 -2.9
Ayaş 0.2 57.7 54.4 -3.3
Gölbaşı 2.3 56 52.7 -3.3
Polatlı 2.3 57.9 54.3 -3.6
Kalecik 0.2 70.3 66.2 -4.1

WHAT WAS THE CONTRIBUTION OF THE MHP?

Between the June and November parliamentary elections, the Nationalist Action Party (MHP) hemorrhaged votes. After June, party leader Devlet Bahçeli refused to form a coalition with the Justice and Development Party (AKP); yet, as violence worsened during the following months, the AKP and Erdoğan were able to present themselves as the defenders of the Turkish nation, thereby co-opting the MHP’s raison d’être.

The loss helped galvanize internal opposition to Bahçeli and the party leadership, miring the leadership in a struggle to maintain control. The coup attempt aided Bahçeli and the leadership’s effort to paint their opponents as Gülenists.[15] Moreover, by fulsomely supporting Erdoğan’s efforts to expand his presidential powers, the MHP leadership has managed to avoid the repression that the Peoples’ Democratic Party (HDP) and even Republican People’s Party (CHP) are experiencing.

The question is whether the MHP leadership convinced those Turkish citizens who would typically vote for it. The results would suggest not: if we look at the twenty-one provinces where the MHP won more than 15% of the vote in November 2015, it appears that nearly 70% of the party’s voters may have rejected the stance of Bahçeli and the leadership.[16] Though the MHP contribution may have made the crucial difference in the referendum’s victory, the nature of the outcome creates an opportunity for the party’s internal opposition in the coming months.

The table below compares the MHP vote share in November 2015 with the difference between the AKP vote and the referendum YES vote. This difference (I am assuming) consists of voters who do not like the AKP, but like the idea of a strong executive defending the Turkish-Islamic essence of Turkey (i.e. MHP voters). By calculating these YES voters as a percentage of the MHP’s 2015 vote, we get some idea of how many MHP voters followed the official party-line.

MHP Leadership Effect (Results in Provinces with +15% in 2014)
November 2015 Change in AKP 2015 to 2017 % of MHP Voter Defection (?)
Adana 19.6 5 75
Afyonkarahisar 17 1.2 93
Aksaray 18.1 3.9 78
Amasya 20.6 5 76
Antalya 17.6 -0.4
Aydin 15.8 1.9 88
Burdur 19.7 1.2 94
Çanakkale 15.6 -0.3
Çankırı 21.6 4.4 80
Isparta 20.4 3.2 84
Karabük 19.8 -0.4
Karaman 16.2 -0.7
Kayseri 18.4 2.2 88
Kırıkkale 20.9 0.1 99
Kırşehir 24.3 2.6 89
Kilis 18.4 -1.5
Mersin 21.4 5 77
Nevşehir 18.4 2.8 85
Niğde 18.8 2.7 86
Osmaniye 34.4 11.3 67
Uşak 18.2 0.4 98

SO . . .WHAT?

Overall, the YES vote was 3.7 percentage points higher than the vote for the Justice and Development Party (AKP) in its November 2014 election. At the same time, the YES vote was 0.4 points lower than in the 2014 presidential election. Though Erdoğan achieved his goal of winning the election and securing additional powers, this feat may be pyrrhic: it was achieved with the help of an MHP leadership that has been shown to be out of touch with its own voters; it was achieved with diminished support in key regions like the Black Sea; it was achieved with increased support in southeast regions where tens of thousands of citizens did not reach the voting booth.

President Erdoğan has successfully cobbled together a coalition of voters sizeable enough to win, but governing will be the greater challenge.

– – – – – – – – – – – – – – – – – – – – – – – – – – –

[1] A version of the Selçuk Tekay song can be heard HERE, sung by Muazzez Abaci (Cengiz Semercioğlu, “Selçuk Tekay, Başbakan Erdoğan’dan telif istiyor…Hürriyet, 1/21/14).

[2] In his victory speech, President Erdoğan made sure to thank the Great Union Party, an extreme nationalist splinter-faction of the MHP, and HÜDA PAR. HÜDA PAR is the shortened name of the Free Cause Party (Hür Dava Partisi) but “hüda” also means “God” (as in “Party of God”) and, sure enough, it is closely linked with the Turkish group Hezbollah (“Party of God”), the Islamist Kurdish answer to the Marxist Kurdish PKK. Over the last two decades, Hezbollah has been responsible for many deaths in its efforts to target left-wing Kurdish nationalists. HÜDA PAR was launched in 2013 and, clearly, is in the government’s good graces (Kadri Gürsel, “New ‘Party of God’ Will Divide Kurdish, Turkish Islamists,” Al Monitor, 12/23/12; “Huda-Par’s emergence,” The Economist, 11/23/13; Jenny White, “What is Huda-Par?” Kamil Paşa, 12/28/14)

[3] A spate of articles giving an overview of conditions in Turkey have come out, timed to coincide with the election. Among the most interesting are Suzy Hansen, “Inside Turkey’s Purge,” The New York Times Magazine, 4/13/17; Mark Lowen, “Erdogan’s Turkey,” BBCNews, 4/13/17; “The vote that will determine the fate of Turkey’s democracy,” The Economist, 4/15/17; Guney Işıkara, Alp Kayserilioğlu, and Max Zirngast, “Voting on Dictatorship,” Jacobin, 4/15/17.

[4] Even dictatorships have elections, political rallies, and parliamentary institutions. They may serve different functions, but they are still worth considering. For example, Jennifer Gandhi sees parliaments and political parties as means by which dictators negotiate with different elements of society while maintaining their own power; Barbara Geddes suggests elections are a means of mobilizing the population in order to discourage challenges to the dictator from the military; and as for rallies, Mariela Szwarcberg hypothesizes these are a way for dictators to evaluate the abilities of their subordinates to mobilize the population (Barbara Geddes, “Why Parties and Elections in Authoritarian Regimes?” Annual Meeting of the American Political Science Association [2005]: 456-471 Jennifer Gandhi, Political Institutions under Dictatorship [Cambridge University Press, 2008]; Mariela Szwarcberg, “Political parties and rallies in Latin America,” Part Politics 20, no. 3 [2014]: 456-66

[5] Human Rights Watch puts the number of displaced persons in the southeast at 400,000; the International Crisis Group puts the number at 300,000 (“World Report 2017: Turkey,” Human Rights Watch; “Turkey’s Refugee Crisis: The Politics of Permanence,” International Crisis Group, 11/30/16).

[6] In 2017, turnout in Şırnak was 83.9% (214,666 votes); in November 2015, turnout was 90.3% (225,228); in June 2015, turnout was 92.1% (228,855); and in 2014, turnout was 82.1% (193,390). It should also be noted, however, that around 30,000 more voters supported YES than had voted for either Erdoğan or the AKP in recent elections—which does suggest that the HDP coalition was less cohesive than the party’s leadership might have hoped. For Şırnak results see Sabah 2014; 2015; and 2017.

[7]Adaylar Oy Dağılımı: Istanbul,” Hürriyet.

[8]İstanbul 1 Kasım 2015 Genel Seçim Sonuçları,” Haberler.

[9]Referandum2017: İstanbul,” Hürriyet.

[10] This analysis ignores the fact that turnout rose from 77% to 88% between 2014 and 2017—and, therefore, the very real possibility that voters who voted for Erdoğan in 2014 also voted for his expanded powers but were simply outnumbered by an increased turnout from anti-Erdoğan voters.

I cannot pretend to have an answer to this, but further research might be worthwhile. A good test case would be a municipality like Ümraniye, where the YES vote was 0.3 percentage points lower than the November 2015 AKP vote. The municipality gave Erdoğan 56.3% in the three-way 2014 presidential race and gave the AKP 55.5% in November 2015. It passed the referendum with 55.2%. In 2017, turnout in Ümraniye was 89.5% and 239,455 voted YES. In 2015, turnout was 88.8% and 235,049. So, although 4,000 more people voted for YES than the for AKP, that is far less than a 0.7% increase in turnout would predict (not to mention the population growth over two years).

[11]Ankara 1 Kasım 2015 Genel Seçim Sonuçları,” Haberler.

[12]Referandum2017: Ankara,” Hürriyet.

[13]Adaylar Oy Dağılımı: Ankara,” Hürriyet.

[14] Istanbul and Ankara population statistics from the Turkish Statistical Institute (TÜİK).

[15]Meral Akşener’den itiraf gibi ‘FETÖ’ açıklaması,” Akşam, 12/3/15; “İşte Meral Akşener ve FETÖ ilişkisi!Sabah, 4/23/16; “Meral Akşener’in Ekibi FETÖ Bağlantılı Çıktı!” Haberler, 9/16/16.

[16] In November 2015, the AKP won 49.5% of the vote nationwide; the MHP won 11.9%. In theory, the combination would suggest a 61% YES vote. Yet YES received only 51.4%, suggesting that the MHP contributed only 2%—enough to push the referendum to victory (good for Erdoğan), but suggesting only 16% of MHP voters went along with the leadership (bad for Bahçeli). Smarter people than me put the number at 35% of MHP voters followed the leadership (“Gezici Araştırma’nın sahibi Murat Gezici AKP’nin oylarını açıkladı,” Cumhuriyet, 4/17/17). Obviously, these are generalizations which don’t take into account HDP and CHP voters who might have voted for increased presidential powers. It also treats the AKP vote more cohesively than it is in reality. However, even if we double the contribution of MHP voters, it would still be a failure for the MHP leadership—only less abject.

[Picture at top from CNN, used for non-commercial purposes.]

Advertisements