Undecided, floating votes to determine referendum results

Undecided, floating votes to determine referendum results

Undecided, floating votes to determine referendum results The outcome of the April 16 referendum on constitutional changes will be determined by the undecided, as has been the case in previous elections, according to a prominent pollster.

More than 80 percent of the electorate will vote according to party affiliation, said Adil Gür, adding that 15 to 18 percent of the electorate is currently undecided. 

“Even in the elections with the highest rate of participation, around 10 percent do not go to the ballot box. So the real undecided are around 10 percent,” he said. That undecided group is made up of floating votes without party affiliation, added Gür, the head of the A&G polling group.

It seems there is a big degree of mistrust in polls lately; what’s your reaction?

The mistrust should not be toward the polls but toward the electorate; because not only in Turkey but the world as well, it is the electorate that takes us by surprise. 
Voters sometimes pretend and prefer to hide their vote. 

In addition, expectations have risen about opinion polls. We said for instance that Recep Tayyip Erdoğan would be elected president with 54 percent of the votes in 2014 elections, he got 52 percent and we were blamed for a mistaken prediction! After all, the polls are not the art of knowing; it shows certain trends.

As we are heading toward a referendum, do you think there are some who are hiding their true colors?

It is a possibility; after all, there is a state of emergency. But this aspect is being exaggerated by the public. Neighborhood pressure changes according to regions. A supporter of the AKP [Justice and Development Party] could hide his color in İzmir, and the same can be valid for a CHP [Republican People’s Party] supporter in Konya. I do not think that the ratio of those hiding their true intentions are that big – it remains within the margin of error.

But it seems the “yes” and “no” are neck and neck, so it might make a difference this time.

That would have been true if we were a week ahead of the referendum. We are two months away from that date. It would be inappropriate to make a prediction today about the outcome.

 What does it look like currently?

The “no” votes were ahead of the “yes” votes even before the constitutional changes came to the parliament. Today, the “yes” votes are a step ahead of the “no” votes when you distribute the undecided. There are at least 15 and sometimes that goes to 18 percent who say, “I am undecided, I will not go to the ballot box or I will give a blank vote.” But as I said, it is too early to make a prediction since the campaigns have not started yet.

Would you say that the undecided are more numerous in comparison to past elections?

No, on the contrary, they are smaller. More than 80 percent of the electorate who are voting according to party affiliation and who have been polarized consciously by politicians know where they will cast their votes. It is impossible for someone who decided to vote “yes” or “no” a month ago to change their vote even if you were to run the most effective campaign. The electorate in Turkey decides according to party affiliation and the result of all elections are determined by the undecided. While the undecided currently appear to be between 15 to 18 percent, I think the real ratio of undecided is around 10 percent. When you ask people how they will cast their vote, instead of saying, “I will not go to the ballot box,” they say they are undecided. Even in elections with the highest participation rate, it is not more than 90 percent. Ten percent of the electorate doesn’t go to the ballot box.

So we should assume that half of those who are undecided will not go to the ballot box. 

But why do you say the undecided are smaller this time?

Because the issue is very simple. It is “yes” or “no.” I don’t mean to say the voter knows every issue. On the contrary, voters do not know and do not want to know. They act according to their party’s line. Neither those who say “yes” nor those who say “no” question the issues. The convincing will be done to the undecided that are probably less than 10 percent.

And they the ones without party affiliation?

Indeed, otherwise, how could we explain the difference between the June 7 and Nov. 1 elections in 2015? What changed in Turkey which made 10 to 15 percent of the electorate change their votes?

The floating votes had gone away from the AKP in the June 7 elections, whereas on Nov. 1, the floating votes went to AKP. They said, “A government wasn’t formed after the June 7 elections; I have a lot of difficulties, I need a stable government.”

But why do you think the undecided would make a difference this time, too? After all, the “yes” camp is made up of AKP and MHP [Nationalist Movement Party] and the number of their electorate combined should be enough to secure the outcome.

All of the electorate that voted for one party is not fully ideologically loyal to that party. In every party there are floating votes.

Between the June 7 elections and Nov. 1 elections, the voters changed places:  The MHP and HDP [Peoples’ Democratic Party] lost votes, the AKP’s votes increased and the CHP got the same amount. The AKP’s votes increased 9 percent; the MHP lost 2 percent and the HDP 1 percent; well then, where did this 6 percent come from? There are deep currents below. In each election, at least 20 percent of the electorate change their places; but we do not see them. We just say the CHP, which got 25 percent, got 25 percent and we think everyone stayed where they were. But that is not the case; there are many people coming and going.

The undecided are important because on Nov. 1, the AKP could have formed the government alone had it gotten 49 percent of the votes or 45 percent of the votes. This time the competition is neck and neck, so 1 percent could make a difference.

So you admit that there might be a neck-and-neck race in the referendum.

Up until July 15 [2016], the “no” votes were ahead of the “yes” votes in all opinion polls. With the failed coup of July 15, the perception has started to change and in all the polls conducted afterward, the rate of the undecided has started to change, and the “yes” votes have started to go ahead of the “no” votes. There are two reasons for that: the number of those who started to say, “It is better to have a presidential system to fight against coup plotters and terrorists,” went up. Second, the confidence toward President Recep Tayyip Erdoğan has become higher. Even today, the support toward Erdoğan is higher than the support toward the AKP or “yes” votes.

So the campaigns are going to be important. What will matter to the floating undecided votes?

Campaigns will obviously make a difference. If this will happen, that will be very bad: One side will create the perception that those who say “no” will be siding with the terrorists, and on the other side, the no-camp say that if there is a “yes” outcome there will be a regime change and the republic will be destroyed.
Campaigns based on the fear factor will be counter-productive. Everywhere in the world, people vote for hope.

It will then be difficult for the “no” camp to have a positive campaign.

It is generally difficult to run a “no” campaign. The “yes” camp always starts 2-0 ahead. That’s the case everywhere in the world.

The economic situation is deteriorating. Will that have an effect on voter behavior?

 I have always argued that people vote in accordance with their economic situation. I said just now that 80 percent vote according to party affiliation, but those who vote in accordance with ideology are lower than 50 percent. These two things are different. In 2009, the AKP lost votes because of the economic crisis. But this time, there is a difference. While there are economic difficulties, the electorate is not holding the government responsible; the perception is that foreign forces are responsible. 

How about the MHP constituency?

The MHP has two ideologically different constituencies. MHP voters in the Mediterranean, Thrace and Aegean are more secular and resemble the CHP constituency. The MHP constituency to the east of Ankara resembles the AKP constituency; they are more nationalist and conservative. In the 2010 referendum, despite the MHP’s “no” campaign, half of MHP voters voted “yes.” In 2014, some MHP voters voted not for their candidate but for Erdoğan. Independent of what [MHP leader Devlet] Bahçeli says or what the opposition within the party say, MHP votes will be divided in the referendum.

But the faith of the referendum will be determined less by the MHP votes but more by the floating votes within the AKP. The AKP received 49.5 percent in the Nov. 1 elections. The AKP and MHP votes combined made 62 percent on Nov. 1, and they stand at 65 percent currently. But there is a shortage among the “yes” votes. So what needs to be convinced are those who voted AKP in the Nov. 1 elections. 

In other words, nearly 100 percent of those who voted AKP in the June 7 elections will say “yes.” The ones who remain undecided within the AKP constituency are the floating votes that went to the AKP in the Nov. 1 elections.

The participation rate will also be important, I assume.

The motivation of the electorate will play a crucial role in the outcome. To what degree the parties will succeed in getting their voters to go to the ballot box will be important. I think there will be a high rate of participation, especially because of the referendum date. Schools are open, people are relatively less motivated to go for a picnic or visit a relative during that weekend.

Who is Adil Gür?


Undecided, floating votes to determine referendum results


Adil Gür was born in 1965 in the Eastern Anatolian city of Malatya. He went to elementary and high school in Malatya and went to Istanbul for university. He graduated from the Istanbul Law faculty.

He became familiar with the research sector during his university years. Between 1987-1995, he worked for Konda Research and Consultancy, which was then part of the Doğan group. 

Adil Gür then set up A&G Research in 1995 with a group of researchers that currently have 30 years of experience in the sector. The company first gained prominence with its accurate prediction in the 1999 elections predicting a loss of vote in the ruling party in contrast to the estimations of other polling agencies.
Adil Gür is a frequent commentator on tv channels.