The Kapali Carsi

Reader Appreciation Day

Loyal readers from the old blog would know that I regularly do reader appreciation days. Reader “medic” had a couple of questions at the comments section of the my latest Hurriyet Daily News column, which was also posted here, so I would like to take those on.

Her first question is about Turkey’s trade balance and the lira:

Today, Turkey imports much more than it exports. How does that affect Turkey’s economy and the Lira in the near and far future?

I see the relationship more going from lira to the trade balance than the other way around. But even than, the relationship is less clear-cut than you’d think. Exporters and the government are quick to blame the deficit on the strong lira, which is not so strong nowadays anyway, but a recent CBT paper, based on survey results, found that the very same exporters choose intermediate inputs because the Turkish counterparts are not as good.

As how it affects Turkey’s economy: It has become an integral part of Turkey’s foreign-financing based unsustainable growth model. In the short run, tight fiscal policy would provide relief, but in the long run, we’d need structural reforms that would make Turkish industry, not just exporters, more competitive and induce Turks to save more.

Medic is also wondering about Turkish statistics:

How is Turkey’s statistics figures of the economy, unemployment, trade etc. collected? Can we trust those figures?

I would say, yes. I have several weird stories about my experiences with TurkStat, from their refusing to release survey questions for fear that people would conduct surveys in their name (I am serious), to not releasing city codes in household surveys. And they have messed up quite badly a few times, the most notable being failure to let the Central Bank know that they were changing the weights to the inflation indices in 2009, resulting in the Bank to totally miss its inflation forecast.

But I would trust the statistics, bearing in mind that no statistic is perfect. Trade figures are rather accurate, as they are registered (excluding drugs, etc. of course), and unemployment comes from a survey. The reason I called unemployment “one of the least reliable of Turkish macro statistics” is not because of the way TurkStat gathers data, but because there is a huge informal sector in Turkey. Because of all those unregistered workers, payrolls data (as in the U.S.) are not reliable, and you have to get all the info from the surveys. Similarly, many firms pay low “official wages” and hand the rest to their workers in cash. As a result, I am surprised that the CBT would use wage data as an argument in the MPC one-pager.

Comments are closed.