Monday, January 30, 2012


YUSUF KANLI Yusuf-Kanli

Polls in the country of rabbits

HDN | 5/30/2011 12:00:00 AM | YUSUF KANLI
With opinion poll companies predicting wildly different numbers in terms of voter support for Turkey's political parties, it is anyone's guess as to the outcome of the June 12 polls.
Less than two weeks before the June 12 parliamentary elections, there is a glossy public opinion poll on my desk. Like many people, I do have an allergy against public opinion polls because of the bad reputation that whoever pays for them is presented with the results that would most please them. Of course that does not mean all the public opinion polls are crooked or faulty. Definitely there are companies in Turkey doing public opinion polls through scientific means and indeed objectively.
Another problem is, of course, the sudden increase in the number of eligible registered voters. This issue, which somehow escaped attention of all of us, was brought to the forefront by Bülent Tanla, a former politician and a pioneer of public opinion polls in Turkey. Only last year at the Sept. 12 referendum, the number of eligible registered voters was around 49 million and in 2007 it was 42 million and in 2002 it was 41.4 million. In 2010 and 2011, it all of a sudden reached 49 and 52 million, respectively. How? Are Turks multiplying like rabbits? Particularly, how have Turks multiplied by three additional million since September 2010, resulting in the number of eligible voters increasing from 49 million to 52 million? What has happened? Or, has someone placed in his pocket in advance some 10 percent of the vote in case of any emergency? It smells bad, does it not?
My problem with the public opinion polls is that so far I could not come up with a public opinion research company in this country who successfully and accurately forecasted three consecutive general or local elections. For every election there is generally a public opinion company, which came very close in its reports to the actual outcome of the poll. And, of course there is nothing surprising in that as there are companies suggesting that the ruling Justice and Development Party, or AKP, will have an electoral support as low as 36 percent in the June 12 election, while there are companies claiming the ruling party will do very well and get support of some 52 percent of the electorate. If the AKP receives between 35 and 54 percent of the vote, then there will be companies declaring victory and celebrating their accurate forecast on the June 12 evening. For God sake, there is almost 20 percentage points between the lowest and the highest forecasts, how scientific can such forecasts be?
Or, the main opposition Republican People’s Party, or CHP, is tipped to receive around as low as 23 percent and as high as 35 percent in various public opinion polls. The fluctuation between the lowest and the highest forecasts as regards the CHP vote, thank God, is a “more reasonable” 12 percentage points.
Well, if in this country there are 52 million registered voters, every single percentage point of support for a party means around 520,000 votes cast for that party. How popular support for a political party fluctuates, if we take the estimates for AKP vote as an example, with a margin of almost 10 million people. Or we are to take the estimations for the CHP vote, the main opposition party may at least get vote of some 12 million people, but may as well receive the support of 18 million people. These companies must have been joking with the intellect of the Turkish society. With such huge margins, of course at least one company will have the probability of declaring success once the vote count is completed on June 12 evening.
The poll on my desk is somewhat different from other public opinion polls as it was not sponsored by any Turkish political party. It was done for the International Republican Institute by the Infakto Research Workshop in collaboration with Harris Interactive.
According to the poll, the AKP has popular support of around 40 percent while the support for the main opposition CHP is around 21 percent, or just slightly more than the half of the support for the AKP. The Nationalist Movement Party, or MHP, which for the past few weeks has been battling to save its reputation from a series of sex-tape scandals, appears to have the support of just 10 percent of the respondents of the poll, just enough to beat the anti-democratic 10 percent national electoral threshold required to be considered eligible to send deputies to parliament.
The Peace and Democracy Party, or BDP, or the political wing of the separatist Kurdistan Workers’ Party, or PKK, terrorist gang, comes next in the popularity list with 5 percent support. The BDP, anyhow, like what it did in 2007 elections, is not participating in the June 12 elections as a political party either. Instead it is participating in the election with a foray of independent candidates seeking election to Parliament from many southeastern, mainly Kurdish-populated, cities as well as some other cities elsewhere in the country, including Istanbul. Last time the independents supported by the political wing of the PKK produced 23 deputies. In this election their target is to send Parliament at least 35 deputies and many political analysts believe they have the probability of winning at least 30 parliamentary seats.
The IRI poll was conducted in the second half of April, before the series of sex-tapes scandals, which killed the political prospects of some 10 leading candidates of the nationalist MHP. Thus neither probable image erosion, nor a possible backlash in the form of boost in the electoral support of the party could be seen in those forecasts.
Yet, since the IRI poll did not distribute the around 23 percent “undecided” or “did not know” vote to the parties, it might be said, according to this poll, the AKP vote might be around 45 percent, the CHP vote around 26 percent and the MHP with a clear 12-13 percent, which would put them over the threshold.
As they say for fortune tellers, don’t believe it, but it is nice having it.

No comments: