Can Turkish democracy count on defected conservatives, rather than Kurds?

Polls conducted over the last few years on Turkish society's sociological profile reveal that at least 60 percent of the population is made up of pious, conservative and nationalist people. This largely corresponds to the totality of votes that the ruling Justice and Development Party (AKP) and the Nationalist Movement Party (MHP) got in the last general election in November 2015, with the AKP winning around 50 percent and the MHP winning around 10 percent of the votes.

The "yes" camp in the April 16 referendum on shifting to an executive presidential system made up of the AKP and the MHP was expecting at least 55 percent of the votes. But despite an unfair campaign period that took place under a state of emergency, "yes" votes remained only around 51 percent.

Some estimate that around five million conservative votes defected from the "yes" camp. There is little surprise about the scale of the defection from the MHP leadership. It looks like around 30 percent of the MHP's constituency voted for "yes" in the end, not out of loyalty to the MHP but out of sympathy for President Recep Tayyip Erdoğan. These are largely the pious conservatives living in Central and Eastern Anatolia. They for example voted for Erdoğan during the presidential election of 2014 instead of the joint candidate supported by the MHP and the main opposition Republican People's Party (CHP). 

However, it seems that conservative nationalist MHP voters in the big urban centers, as well as those living in Aegean and Mediterranean coastal regions, largely voted "no." Did they vote "no" because they were against the presidential system, or because they were angry at MHP chair Devlet Bahçeli's stance against dissent within the party? Was it a reaction against economic problems,...