All Politics is Local, 2015: Ill Tidings for the Turkish Opposition
Provinces matter in Turkish elections. Understanding why is crucial for appreciating the scale of the Justice and Development Party (AKP) victory on November 1 and the true significance of the elections.
Turkey has eighty-five electoral districts—essentially one for each of the country’s eight-one provinces with the largest three provinces (Istanbul, Ankara, and Izmir) split into multiple voting districts. These three highly urbanized provinces contain nearly thirty percent of the country’s 77 million people, leaving the rest of the country sparsely populated by comparison. Whereas Istanbul accounts for 18% of the population, the province of Çanakkale (fortieth most populated out of the eighty-one provinces) contains 0.7%, and Bayburt (the smallest) just 0.1%. Yet even little Bayburt gets two representatives in parliament. The result is that each of Istanbul’s eighty-eight parliamentarians represents about 200,000 people while each of Bayburt’s represents just 40,000. Yet their votes are weighted equally.
Even these details do not capture the degree to which the electoral apportionment system advantages certain parties over others. The province of Iğdır, serves as a clear example. With a population of 192,000, Iğdır’s allotment of two seats is slightly greater than it would receive through a more equitable distribution. On November 1, the HDP garnered 52% of the vote, giving it one seat. Since Iğdır has a population around 192,000, this parliamentarian will effectively represent 96,000 Turkish citizens. The remaining 48% of the vote was split: 30% went to the AKP, which maintains a similar level of support in most eastern provinces, and 13% went to the Nationalist Action Party (MHP), whose anti-communist nationalism retains popularity near the Russian/Central Asian border. As the second-place party, the AKP won the other seat. In other words, in a province with slightly too many seats, the AKP won a seat with almost half as many votes as the People’s Democratic Party (HDP). The 30,000 votes the AKP earned in Iğdır won it a seat; in Istanbul’s first district, it took parties more than 100,000 votes to win a seat.
The electoral calculus dramatically skews elections toward parties with strength in the countryside. If not for this dynamic, the AKP’s 49.5% of the vote would have given it 272 representatives rather than the 317 it won: four seats shy of a majority.
Until the June 2015 election, the AKP had also benefitted from the 10% threshold that parties must pass nationally to enter parliament (and receive campaign funding from the central government). If a party fails to pass the threshold, its votes are simply distributed among the parties that succeed. Rather than risk such an outcome, politicians representing Kurdish voters in the east had previously campaigned as independents, which allowed them to enter parliament by victory at the provincial level. There was, however, a catch: in a province like Mardin with six seats, the AKP received 32% of the vote in 2011 and the “independent candidates” collectively won around 60%. Since each independent ran separately, the seats ended up split 50-50, effectively giving the AKP an extra seat. In a province like Edirne, by contrast, where the also AKP won 30% in 2011 (but against an official party), it won a corresponding 30% of the seats. In June, the HDP won 76% of the vote and took five seats—even in November it still won four seats with 68%.
By crossing the 10% threshold for two elections in a row, the HDP has denied the AKP one important advantage. Yet the November elections served to accentuate the AKP’s other advantage: strength in the lightly-populated-but-numerous central Anatolian provinces. Although the Republican People’s Party (CHP) increased slightly, it did so in the highly populated regions of the country. While both the AKP and CHP picked off HDP and MHP voters, the AKP was more successful in the areas where it mattered most.
STATISTICS AND ANALYSIS
Key Observation: The MHP declined by double-digits in six provinces. On the “bright” side, some of the largest losses simply negated earlier increases (for example: Aksaray, Arvin, Kayseri, Kütahya, Yalova).The HDP also had large declines, but the party had started from nothing in 2011 and risen incredibly high; by November, it had very few directions to go but down. The MHP, on the other hand, declined in areas of the country where it has generally been strong.
Key Observation: The MHP vote declined from 2011 to June to November in areas that had once been its strongholds (Anatolia and the old Soviet border). In fact, looking back to 2007, we see a three election-cycle decline in Osmaniye and Kars.
Key Observation: The chart speaks for itself.
Key Observation: Though the AKP suffered large declines in June, none of these were repeated in November. In every province, the AKP vote increased. In thirty-two out of the eighty-five voting districts its gain from June to November was in double digits. In the worst June loses (Van, Ağrı), the party recovered nearly half its loses. In many provinces, the November increase raised the party above its 2011 result.
Key Observation: AKP loses in eastern, majority provinces seem to be permanent.
Key Observations: The AKP continues to gain strength in the northeastern Black Sea region, an area where appeals to nationalism have long been popular.
Key Observation: From dramatic highs in June, the HDP lost in every province except Hatay (where it gained 0.2%). Its loses appear most dramatic in the east where the party is the strongest. That said, given the lack of competitive third parties in the region, the actual distribution of seats only changed by one or two in most of these provinces.
Key Observation: The CHP, generally popular among Alevi voters, suffered a large defeat in Tunceli where its fortunes were similar to the AKP: Kurdish voters shifted in large numbers to the HDP in June. Unlike the AKP, however, the CHP did not recover much lost ground in November. Here and there, the party picked up a percentage or two from the HDP and MHP.
Key Observation: Turkey’s Kurdish voters are increasingly split between the AKP and HDP.
Key Observation: In western provinces (Bilecik, Duzce, and Edirne) the CHP benefited first from the AKP’s collapse in June and then the MHP/HDP’s in November. (In Bilecek and Edirne, for example, the change in the HDP vote was too small to explain the increase in the CHP vote.)
* * * * * * * * * * * * * * * * * * * * * * * * * *
 With a population of 77,000,000 and 550 seats, there should be around one representative for every 140,000 people. Iğdır should, consequently have about 1.4 seats. (NB. I’m ignoring the fact that voters are different from population. Were we to consider this distinction, we should note that the HDP received 49,000 votes in Iğdır, which translated to 52% of the total vote. By contrast, 52% of the population would be 99,000.)
 If the parliamentary seats were not calculated province-by-province, the CHP would have gained around five seats and the HDP would have remained the same. The MHP is the party that suffers the most from the system: although it earned more votes than the HDP, it received nineteen fewer seats (40 to the HDP’s 59). If seats were proportional, the MHP would have received sixty-three, four more than the HDP!