The Washington PostDemocracy Dies in Darkness

Why the results of Turkey’s election are surprising

On June 25, the day after Turkey’s election, people in Istanbul walk past a billboard showing President Recep Tayyip Erdogan. (Emrah Gurel/AP)

Last month, an election in Turkey kept President Recep Tayyip Erdogan and his coalition in power. But experts are puzzled by the results — and caution that the election was not free and fair.

Videos of ballot stuffing — mostly in eastern Turkey — in favor of pro-Erdogan parties went viral after they were posted online on election day. And both partisan and nonpartisan reports showed that allegations of electoral irregularity came primarily from eastern Turkey. An opposition-written report stated that 68 percent of the election day violations took place in the east — areas where Erdogan’s Justice and Development Party (AKP) and the far-right Nationalist Movement Party (MHP) experienced significant gains. A report published by an independent fact-checking organization largely supports these claims.

Because this area has a majority-Kurdish population, election observers were surprised at the increased support for pro-Erdogan nationalist parties that suffered significant losses in the rest of the country, despite years of tensions between Erdogan’s government and Kurds.

To sort out the competing claims by pro-government and opposition sources, I examined data at the ballot-box level. I found a significant level of overlap between the geographical distribution of this unprecedented electoral result and that of some electoral anomalies. These include: excess votes per boxes, high levels of spoiled ballots, and noticeable differences in the amount of parliamentary and presidential ballots in the same box.

Excess votes

Election officials were allowed to vote in a different precinct if they were working away from their own on election day — and so analysts were not surprised to see some ballot boxes having more votes than their precinct had voters. But what drew attention is that 42 percent of 3,278 such boxes were located in the predominantly Kurdish east, with consistently pro-government results, although overall only 17 percent of the boxes were dedicated to the east.

Within the districts where the excess vote count at the ballot-box level is more than 30 ballots, the pro-government coalition gained an average of 2,881 new votes. In the remaining districts, on average, it lost 51 votes compared with 2015.

These results become even more striking when we consider the following two points. First, a box with excess votes means a turnout rate at least 14 percent more compared with an average box, because the nationwide turnout was 86 percent. Second, before the election, numerous local governors in the east decided to move hundreds of ballot boxes — with 144,000 citizens registered to vote — from rural areas to district centers. Obviously, this decision had a potential to lower the overall turnout in the east. Hence, we should have expected even fewer boxes with excess votes in the east compared with the rest.

Invalid votes

Across the country, an average of 2.4 percent of the total votes were rendered invalid. But those districts where the pro-Kurdish opposition party had won more than 60 percent of the votes in 2015 had an average of 3.6 percent invalid votes. Pro-Kurdish party officials claim that this outcome is due to the ballot-box officials’ systematic bias against their party and systematic pressure against their ballot-box representatives.

But in the 457 ballot boxes in the east where Erdogan received more than 99 percent of the valid votes, the percentage of invalid votes turned out to be a record low: 0.5 percent. More than half of these boxes were in Sanliurfa province, where the ballot-stuffing allegations were at their peak.

Presidential vs. parliamentary ballots

The total number of parliamentary ballots should — under normal circumstances — equal the total number of presidential ballots in each box. The rule was “two ballots into one envelope.” If the difference between these two is simply because of human error — instead of careless unbalanced ballot stuffing — we should expect no significant variation across regions.

However, I found that 12 percent of the boxes in eastern Turkey have different numbers of each ballot type, whereas only 6 percent of the boxes in the rest of the country have such discrepancies. Possibly, individuals who committed ballot stuffing (allegedly mostly in the east) were not extremely careful in adding equal numbers of ballots.

Fraud allegations vs. shifts in voter behavior

How can we tell this is fraud and not ordinary shifts in voter behavior? And why does it matter?

The opposition has claimed that the results indicate fraud. A pro-Kurdish opposition party published an extensive report that said thousands of party ballot-box representatives and observers were intimidated, threatened or even detained (over terrorism allegations). The report also noted several incidences of ballot stuffing. In one such case, a group of 200 armed men — including a legislative candidate from the southeastern province (Sanliurfa) — allegedly cast numerous illegal votes in various voting stations. After the election, members of the main opposition party said the opposition did not have representatives to monitor thousands of boxes and hence failed to receive ballot-box reports on them.

There are other possibilities. Some election analysts argue that the unprecedented electoral gains of the nationalist parties in the east could be the result of an increase in Turkish security personnel appointed to Kurdish regions during the past few years. However, reports show that the number of security personnel in these regions increased by only about 20,000, which can explain only a very small portion of 596,045 new votes that the pro-government nationalist parties gained in the east. Further, the opposition parties are even more popular than these two parties among military personnel and their families, as the election results from the military housing sites demonstrate.

Others argue that the alliance between the ultranationalist MHP and local tribes may be the main factor. However, the election results show that the MHP’s total vote count increased by more than 20 percent in more than 100 districts in the east. There is not yet any publicly available evidence for the claimed electoral impact of tribes, or of how their influence extended over so many districts.

Looking ahead, the dissemination of videos purporting electoral fraud will erode voters’ hope for political change via the electoral process. The opposition parties’ inability to investigate the extent of electoral irregularities during this election disappointed opposition segments. This disappointment could lead to a substantive decline in turnout in the next elections, ironically only further consolidating the AKP’s power.

Abdullah Aydogan is a research scholar at Rice University’s Baker Institute for Public Policy. Follow him on Twitter at @abdaydgn