Thursday, April 30, 2015

Unpacking Kazakhstan's Election Data

Comrades,

Now that I am back in DC, I can sit down and analyze the various facts, figures and notes I collected over the past few days in Astana.

While it is no surprise that President Nazarbayev was reelected with an overwhelming majority, the political implications of the elections, and the factors driving them in the first place, are more dynamic than have been presented in most Western media sources.

I am no Nate Silver, but as a political-economist, I do love some data analysis. Let’s look at the numbers. The Central Election Commission (CEC) reported that President Nazarbayev was reelected with 97% of the vote, with 95% voter turnout. Kussainov secured .68% and Syzdykov secured 1.68%.

I tallied up the votes from the oblasts reported by the CEC (available here) and I obtained slightly different numbers. As you see, I obtained a voter turnout of 94.93%. The sum of the reported oblast-level data totaled 9,036,724, whereas the CEC reported 9,090,920 voters. 

This was fishy. So I wrote to my contact at the CEC, who is the head of international affairs for the body. He responded that the difference can be explained by votes that were not counted because they indicated multiple candidates on their ballot and/or it was not possible to determine which candidate the voter selected. In total, there were 54,196 ballots that were discarded because they were illegible. This is understandable. After all, let us not forget the saga of the 2000 Presidential election with the vote count in Florida.




(On a side note – my contact has been very transparent and helpful both in Astana and communicating via email. As a foreigner, my interactions with government officials in any former Soviet state can be very hit or miss, so cheers to him for his professionalism.)

I spoke with the CEC after they announced the votes. In total, 93 complaints were registered. Most of these complaints were focused on issues surrounding the candidates, but no complaints were reported on violations of an individual’s right to vote.

Because the OSCE/ODIHR did not report major fraud, and noted that the elections were efficiently administered, I am going to accept the data as it is presented. Given the high turnout, is it possible that there was some sort of manipulation of votes? Sure. But I do not have the evidence to substantiate such a claim. Plus, in some of my more academic work, I deal with a lot of data from this region, and I can assure you, there is no such thing as good data. You work with what you have.



The numbers are quite revealing.

First, Nazarbayev secured the lowest percentage of the vote in Almaty city. With 92.55%, while this is the highest of all three candidates, it is the lowest level of support out of all of the oblasts and administrative districts.

Second, Kusainov proved to be a marginal candidate. I interviewed the leadership team of Kusainov. They indicated that they set up field offices in almost every oblast, except for oblasts they knew they did not have a chance. One of those oblasts was Mangistau. Not surprisingly, Kusainov only secured 707 votes there. Kusainov performed best in Almaty city (1.48%) and in Karaganda oblast (1.07%). Kusainov is from Karaganda oblast so he is a “favorite son.” In Almaty, voters who did not support Nazarbayev likely selected Kusainov as an alternative.

Third, Syzdykov performed remarkably well in Almaty, with 5.97% of the vote. I suspect that he attracted many older voters. He did less well in Astana city, securing 1.12% of the vote. The average age in Astana is 32 and almost everyone works in the government in some capacity, so it is not surprising that support for Syzdykov was low (but higher than in other oblasts and cities), and that Nazarbayev secured a strong 98.54% of the vote. Interestingly, Syzdykov performed the strongest in Mangistau oblast. Mangistau is an oil-producing region that borders the Caspian. According to Mangistau’s Development Strategy 2011-2015, the oblast had the second highest per capita GDP in Kazakhstan of 2542,5 thousand tenge in 2009. This is also 2.3 times higher than the national average for per capita GDP of 1068 thousand tenge. This is however, an inflated number because of the amount of oil produced and the high salaries of international oil workers. Many low ranking oil workers do not obtain attractive salaries and Mangistau has historically been a place of discontent and unrest. There were riots in the Mangistau in 1989 and in 2010. It is therefore not surprising that voters showed great support for Syzdykov.

On a methodological note, voters are registered to vote by their propiska – or registration. Every citizen in Kazakhstan has a propiska that is associated with an oblast. If I am living in Almaty city, and my propiska is in East Kazakhstan Oblast, in order to vote in Almaty, I must obtain the necessary documentation (like an out of state ballot) to cast the vote. In terms of the results of the data, this means that there are voters who are voting in areas outside of the regions where they are registered. Many people who work in Astana, for example, still have their Almaty propiska, simply because they haven not yet gotten around to the paperwork. Comrades, do not read into this, as this happens all the time in the US. For example, I am a resident of Virginia, but am registered to vote in NJ.


I hope that was a helpful exercise, readers. I have much more commentary to come, but I have received some feedback that my posts tend to be very long, so I am trying to keep them short, to the point and more frequent. I look forward to your feedback!

No comments:

Post a Comment