The column is just an elaboration of Friday's post. There are many more things to say, but as I explain below, I just can't:(
In fact, I don’t have much choice: I recently came to agreement with polling firm Konda to work as a consultant on their monthly Barometer Survey, or KBS. Since my findings and writings will be used in their subscription-only report, I am not supposed to give out anything, and hence this column has to be about nothing.
Other than the standard stuff on respondent characteristics and demographics, the survey asks the respondents on how they, their city and the country have been affected by the crisis as well as their views on the recent political developments such as Sledgehammer, Ergenekon and the like. There are also questions on voting preferences, not only for today but also in the 2007 elections.
These questions present a golden opportunity for an economist to set up a simple statistical framework on the determinants of voting behavior. After all, it is a well-known fact that, as Bill Clinton campaign strategist James Carville so eloquently put, it’s the economy, stupid: Economic conditions, whether they are by pure coincidence or the result of deliberate policy, are the best predictors of elections in advanced economies.
This relationship has tended to hold for Turkey as well, although there are much fewer studies for Turkey than for the U.S. and other advanced economies. A model based on a survey could improve on the existing studies: After all, people do not vote by looking at the latest data, but by how they, their family and friends have been doing as well as their perceptions on the country’s economic conditions.
Of course, sampling problems could lead you astray in a survey. Luckily, KBS also has control questions about verifiable data, and those hint that the sampling has been done well. For example, the percentage of people who lost their job during the crisis, as derived from the survey, is line with the figures in official statistics.
In general, the respondents have a surprisingly clear grasp of the economic realities of the country. There is some evidence that they suffer from money illusion, the tendency of individuals to think in nominal rather than real terms. But a quick literature survey revealed that the size of the KBR money illusion is not more than in other similar surveys. The average Turk is much more of a homo economicus than you (or at least I) would think.
These respondent characteristics would not bode well for the Justice and Development Party, or AKP, in theory, especially combined with the fact that, contrary to the PM’s claims that it passed tangent to Turkey, the crisis has hit the country hard: Wednesday’s GDP release for the last quarter of 2009 is likely to show that the economy contracted around 5.5 percent last year, one of the highest in the world.
Unfortunately I am not allowed to disclose the results of my election model, but suffice it to say that I was rather baffled at first, humbly attributing them to my deficiency in Econometrics. But I did manage to find an answer by creating indices of economic and political polarization.
I wish I could share the graphs of the distribution of voting preferences based on economic and political polarization, which caused quite a bit excitement at Konda Friday evening, but if you are interested in learning more, you will just have to contact them.
By the way, Konda is paying me a fixed consulting fee, not a share of revenues, so I have no incentive in marketing their reports. I am just your friendly neighborhood economist who wanted to find out what writing an economics column about nothing would feel like.
Emre Deliveli is a freelance consultant and columnist for Hurriyet Daily News & Economic Review and Forbes as well as a contributor to Roubini Global Economics. Read his economics blog at http://emredeliveli.blogspot.com.