MESOP FACTS ON TURKEY & TURKEY’S UPCOMING NEW ELECTION – Polling companies vulnerable to manipulation: enabled by parties and media

FACTS ON TURKEY RESEARCH : March 1, 2015 – As Turkey approaches yet another critical election on June 7, the reliability of polling companies is in question, with the companies themselves acknowledging that their sector is vulnerable to manipulation, government pressure and the effect of mediawilling to disseminate distorted results. Amid growing rumors that the governing party is applying pressure on polling companies to announce numbers that reflect it favorably, politicians and polling company executives have refrained from providing names out of fear of persecution in the absence of concrete evidence.However, Sunday’s Zaman pointed to countless examples of likely manipulation. One recently contested example is polling company Konda’s inaccurate estimation of the Aug. 10 presidential election in which President Recep Tayyip Erdoğan seemed to be far ahead of his rival, Ekmeleddin İhsanoğlu. Although publishing poll results within the last 10 days before an election is prohibited, Konda did so, and announced that Erdoğan had 57 percent support, even though the final vote turned out to be 51.7 percent. Konda later apologized for its mistaken findings.

When Sunday’s Zaman contacted Bekir Ağırdır from Konda, he initially refrained from answering questions about allegations of manipulation but later said that Konda had submitted its poll results to the Turkey Researchers Association (TUAD) following his company’s inaccurate findings. However, according to TUAD, as well as other polling company representatives, no independent auditing mechanism for polling companies currently exists in Turkey. “There is no inspection in our sector,” said Professor Özer Sencar, the head of the MetroPOLL polling firm. The director of the Gezici Research Company, Murat Gezici, stated that TUAD has no legal inspection authority and that membership is voluntary. Gezici went on to say that after his company announced poll results showing support for the ruling Justice and Development Party (AK Party) to be below 43 percent, TUAD emailed him asking for the source of his research. When Sunday’s Zaman contacted TUAD about Konda’s inaccurate results, they sent a statement from October 2014 in which they explained the mistake as “sampling variability.” The TUAD report criticized polling companies that did not share their data with them. TUAD did not respond to a question about the violations of the ban on publishing poll results within the last 10 days before an election.

Gezici provided a clear example of the external pressure applied to polling companies. On Feb. 24, shortly after Gezici shared the findings of a recent poll in which the AK Party’s support was at 39 percent, tax inspectors paid an unexpected visit to his office. While this incident was largely interpreted as an attempt to intimidate other polling companies, Gezici said such tactics actually work: Even international companies based in Turkey do not want to work with him when compiling their market research.

After revealing that a mayor of one of the three metropolitan cities in Turkey offered him $200,000 to falsify election polls, Gezici said he is considering a job change. “I urge all polling companies to be transparent about the projects they receive from government institutions,” Gezici said, revealing that a different polling company had recently received a $10 million contract in exchange for favorable results for the AK Party. Konda, which has been under fire from the media and opposition parties, has always denied involvement with government institutions. Ağırdır denies having conducted polls for parties or candidates. He also refused to comment on accusations of deliberate manipulation. According to a recent study conducted by MetroPOLL, 19.4 percent of respondents said their opinion had been influenced by polls that estimated a clear victory for Erdoğan. More critically, 15.8 percent of Republican People’s Party (CHP) and 13.8 percent of Nationalist Movement Party (MHP) voters admitted that they had been influenced by such polls.

According to Professor Sencar from MetroPOLL, manipulations before the elections were an attempt to prevent certain voters from going to the polls, giving them the impression that Erdoğan would win regardless. İstanbul CHP deputy Sezgin Tanrıkulu has no confidence in the reliability of polls that appear in the media and encourages the public to criticize manipulative results. According to him, companies whose data fall beyond standard deviation should simply close. Tanrıkulu also noted that new companies emerge shortly before elections to influence public opinion in favor of the government, as the AK Party is skilled in “perception management.”

Triple mechanism of parties, polling companies and media manipulates results

While 36.5 percent of people say polls have an impact on their voting behavior, 40.3 percent of the respondents said they believe polling companies manipulate results, according to MetroPOLL’s findings in February.Politicians accuse polling companies of distorting results in order to win favor from the government and receive benefits from state funds and media outlets who are willing to publish this data. Indeed, based on Sunday’s Zaman’s research, manipulation appears to be facilitated by a triple mechanism of political parties, polling companies and media outlets that lack ethical values and do not face legal sanctions.In one example of a series of leaked voice recordings that reveals direct intervention in the media by Erdoğan, the then-editor-in-chief of the Habertürk daily, Fatih Altaylı, is heard speaking with the “media commissioner,” who had been authorized by the government to make sure Habertürk remained government-friendly.

In the recording, which was leaked in February 2014, Altaylı discussed a poll conducted by Konsensus in March 2013, saying: “I moved some of the undecided voters and some from the MHP. This manipulated the results.”

Although Altaylı argued that the tapes were doctored, he did admit that there had been requests for a change in the results.Sudden and unlikely changes in results also took place in polls that were published by media outlets. In a March 2012 Konsensus poll asking who the next president should be, 16.9 percent of respondents chose Erdoğan, while 49 percent opted for then-incumbent President Abdullah Gül. However, only two months later, the numbers of the same polling company changed dramatically. Support for Erdoğan’s presidency rose to 41.8 percent while Gül’s declined to 21 percent.

The same company seemed to have reached somewhat accurate results as far as the March 2013 İstanbul local elections were concerned. Konsensus estimated 44 percent of the votes would go to the AK Party and 41 percent to the CHP. The actual results turned out to be 47.9 and 40.1, respectively. Sunday’s Zaman was not able to reach Konsensus for comment.Mansur Yavaş, a politician who ran on the CHP ticket and lost the local elections in Ankara in a tight and contentious race on March 30, argues that he witnessed manipulation in the polls not only in the 2014 elections but also in 2009. Yavaş had previously appealed to the Supreme Election Board (YSK), asking for a recount of the votes in Ankara due to allegations of fraud. According to him, it is the YSK’s responsibility to prevent voting manipulation, but his appeal was denied.

In the 2014 elections, Yavaş lost the elections with 43.8 percent of the votes as opposed to AK Party candidate Melih Gökçek’s 44.9 percent, according to official figures. Yavaş provides examples from his own experience and says that in a TV show on Habertürk in 2009, Konsensus stated that he had received only 16 percent of the vote when in fact he had received 27 percent. “Many people came to me after the elections and said that they did not vote for me in order not to waste their vote thinking that I had no chance of winning,” Yavaş told Sunday’s Zaman. According to him, some polling companies receive contracts from municipalities in return for the manipulation of polls. “I name some of them. They said that they’ll sue me, but nothing has happened so far,” Yavaş noted.

Apologies do not excuse results

Yavaş is resentful toward “apologies” from the polling companies, arguing that they do not excuse doctored results. According to him, in matters that affect the future of a country, polling companies and the media should be more conscientious and sanctions should be applied. He has concluded that, in Turkey, public opinion polls are used as a tool to make money. When asked whether the companies care about having a reputation for reliability, Yavaş replied: “In such a case, high amounts of money should be paid.”

The media seem to be an indispensable tool to manipulate public perception. Gezici argues that while dailies such as Hürriyet, Milliyet and Vatan refused to publish his polls in which the AK Party has lower figures than generally declared, the same papers do not mind polls that find lower support for the main opposition CHP. Aykan Erdemir, the CHP’s Bursa deputy, finds huge deviations suspicious as he argues that some polling companies have financial stakes in the government. He urges the media to be more cautious about sharing poll results, arguing that, in the presidential election, Erdoğan’s rival İhsanoğlu was portrayed as unlikely to win. Erdemir explained that such manipulations cannot take place in the West due to higher ethical and democratic standards.Professor Sencar agrees, saying, “In the US or Europe a company that even considers manipulation cannot survive.” In a confession of helplessness, Altan Tan, a deputy from the Peoples’ Democratic Party (HDP), joins others in decrying the prevalence of manipulation in Turkish politics, while adding that there is not much to do except to complain. http://factsonturkey.org/16294/polling-companies-vulnerable-to-manipulation-enabled-by-parties-and-media/