Turkey will say 'yes' in next month's referendum, according to poll

Quite in contrast to my predictions of a tight result at Turkey's constitutional referendum on 12 September, GENAR have released a poll that suggests quite a strong yes vote:

Yes: 56.2%
No: 43.8%

The poll predicts a turnout of 87 percent, which is higher than the last referendum (67 percent) and strikes me as rather high even by Turkey's recent electoral record. Predictably, much of the voting is along party lines: a crushing number of governing AK Party supporters (98.1%) will vote yes, while a similarly huge number of opposition CHP supporters (91.8%) will vote no. 

Murat Yetkin writes in today's Radikal that the Kurdish vote is something Recep Tayyip Erdoğan, prime minister, is depending on to pass his prized reforms. The pro-Kurdish BDP has boycotted the referendum, but more than half of its supporters say they will vote anyway and are more likely to vote yes than not, GENAR's polling suggests. BDP leaders have offered to support a yes vote in exchange for some promises for further reform from the government, but polling like this appears to indicate the BDP holds less influence over Kurdish voters than it likes to believe.

GENAR asked about awareness of the package being put up for referendum: 80.3% said they hadn't read the proposed changes. Of those who had, a narrow majority (23% vs 20%) said they would be voting yes. Those who they had followed debates in the media to some degree (59.7%) were more likely to be voting no.

I'll have more analysis of the referendum package - and that all-important Kurdish vote - in the coming days.

General election voting
GENAR also asked how respondents would vote if there was a general election on Sunday. The headline percentages were (with changes from the last GENAR poll I covered in January):

AK Party: 41.0 (+4.5) [Justice and Development Party, governing, religious conservative]
CHP : 28.0 (+5.1) [Republican People's Party, secular]
MHP : 14.9 (-3.9) [Nationalist Action Party, nationalist]
BDP : 5.1 (-2.0)* [Peace and Democracy Party, pro-Kurdish]
SP : 2.7 (-1.3) [Felicity Party, strongly Islamist]
Others : 8.3 (-2.5) [includes independents]

The changes look quite large, but remember that there is an eight-month gap between this and the previous poll. Support for AK is up; support for the CHP is up by a greater amount, attributable to the rise of Kemal Kılıçdaroğlu.

More interesting is the exodus from the right-wing MHP: almost a third of people who voted MHP at the 2007 general election said they would not vote for them again. A quarter of these floaters said they would now switch to CHP. This reflects a trend in recent GENAR polls: many MHP voters appear to be disenchanted with their party. Only 74 percent said they would support their party's position at the upcoming referendum (as opposed to 98 percent and 92 percetof AK and CHP voters respectively). Turkish voters at large seem to feel the same way: 84 percent said they could never imagine Devlet Bahçeli, MHP leader, becoming prime minister.

How would these latest voting intentions look in parliament? Well, keep in mind that the following is a crude uniform swing, assuming the BDP's 20 MPs run as independents and retain their seats (with changes from the present situation):

AK Party: 259 (-77)
CHP : 177 (+74)
MHP : 94 (+24)
BDP : 20 (NC)

So in parliament this would represent a clear swing from AK to CHP. It would also be coalition territory: AK would be just short of the 276 seats needed to govern alone; CHP+MHP together wouldn't be able to reach this threshold either.

GENAR interviewed 2274 people in 16 Turkish provinces between 31 July and 8 August 2010. The full survey can be found here. * Figures compared with the Democratic Society Party, now banned.

Three things that can lose you a referendum

It's referendum time in Turkey again. On 12 September, 49 million or so Turks will be asked to approve the latest package of substantive changes to the constitution drafted 28 years ago by the country's last military junta. In the best traditions of irony, voting day wil also mark thirty years since the coup that put that junta in power.

The last time Turks were consulted on constitutional change was October 2007, when the headline reform was to the presidency. "We want a president elected by the people," the AK government proclaimed, "and may he hold office for two five-year terms." 69 percent of voters agreed. Changes relating to parliament's voting rules and term in office were bundled into the same package. All said, an easy win.

Three weeks tomorrow, voters will be called back again, but this time it won't be as straightforward for the AK Party. Opinion polls indicate the yes vote is ahead, but only narrowly so, and a large proportion of voters are still undecided. Three reasons help explain why 2007 won't be repeated again.

The first is the Turkish army, which is not a factor in this referendum. In 2007, the General Staff published its notorious "e-coup" online, criticising the government and the threat it posed to the secular state. The gamble back-fired: AK called a snap poll, was returned by an increased majority, and went through a honeymoon period that helped it comfortably win the referendum too. Many AK supporters then were simply those alarmed by the prospect of a military intervention.

This time, there is no stand-off with the military. Aside from a spat surrounding the appointment of one particular general to the post of Land Forces Commander, the government and army have been in full agreement - over the fight against the PKK - and there is no anti-coup sentiment to exploit.

This helps partly explain the second reason why history won't be repeated - that the government's support base is shrinking. 2007 saw a strong government with a strong mandate presiding over a strengthening economy; 2010 brings us a weaker government with a nearly-expired mandate, presiding over an economy out of recession but facing unemployment above 10 percent. Put another way, the people are bored with this government and aren't all that richer than they were three years ago. Besides, they now have a credible alternative.

Kemal Kılıçdaroğlu was elected leader of the opposition CHP in late May. Turkish voters were interested: here at last was a personality to rival Recep Tayyip Erdoğan, prime minister, who spoke colloquially about jobs and public services, and didn't have that aura of elitism that followed the former CHP leader around.

Mr Kılıçdaroğlu has maintained his party's support of a no vote, not entirely out of a compulsive rejection of anything proposed by AK, but also because of rational argument: why, for instance, is a package of such diverse reforms being voted as a whole, rather than as individual clauses? The CHP leader's voice will sway many in the next few weeks.

The third reason why this referendum will be no easy win is that the proposed reforms are hideously complicated. I certainly can't profess to understanding them all yet, and I suspect a lot of the Turkish public is with me. Part of the reason for this is the lack of any headlining reform: in 2007, people were promised the right to elect their own president. Most people understood that. In 2010, people are being promised that senior judges will be appointed in a slightly different way. A marketing dream this is not.

As it stands, the governing AK party is supported by the far-right Great Union Party (BBP) and the religious Felicity Party (SP) in a yes vote. Aside from the CHP, the major parties urging a no vote are the Nationalist Action Party (MHP), the centre-right Democrats (DP) and the centre-left DSP. The pro-Kurdish Peace and Democracy Party (BDP) are currently boycotting the vote, but have incidated they may switch to a yes if some of their demands are met.

But more on that in a future post. Over the coming weeks, I'll examine the arguments of the yes and no camps, cover the political machinations as Mssrs Erdoğan and Kılıçdaroğlu campaign, and come to a - no doubt highly influential - conclusion over what verdict Turks should reach on 12 September.