Sunday, December 22, 2013

Window on Eurasia: Census Data on Nationalities in Daghestan Almost Certainly Falsified



Paul Goble

            Staunton, December 22 – A close examination of the late and still incomplete data from the 2010 census shows that officials have made more than “arithmetic” mistakes in compiling the total numbers and percentages of nationalities in Daghestan and almost certainly falsified the numbers to boost the share of some nationalities and cut that o others.

            On Sp-analytic.ru, Milrad Futulayev notes that the release of Daghestani data on the website of the republic’s statistical committee passed virtually unnoticed a week ago, at least in part because the committee’s officials did everything they could not to call  attention to this information (sp-analytic.ru/news/2730-skrytaya-etnostatistika-ili-pochem-u-tak-dolgo-molchal-dagstat.html).

            There are at least three reasons why Dagestanstat did not trumpet the results. First, its compilation and release of data comes weeks if not months after other federal subjects have issued theirs. Second, it is still incomplete: of the planned eight volumes for Daghestan, only six have been released, although they do include the one on the ethnic composition of the republic.

            And third, and likely most important, there are serious problems with the data. The numbers for one and the same category of people are different in different places and the total number of several nationalities has been changed, with some increasing and others cut (http://dagstat.gks.ru/wps/wcm/connect/rosstat_ts/dagstat/ru/census_and_researching/census/national_census_2010/score_2010/).

            Unlike in other federal subjects, Daghestan has not yet published complete data on the nationality breakdown of the population. But it is possible to glean something about that critical data set by examining table four of the third volume on “Population by Nationality and Langauge by Urban Districts and Municipal Regions of the Republic of Daghetan.”

            Unfortunately, Futulayev points out, this data set is defective: One rural municipal district is missing as are several rural settlements.  Nonetheless, it should be possible to calculate data for those at least in toto by using the summary data provided elsewhere. “But this turns out to be impossible.”
           
            Indeed, if one does that kind of calculation, one discovers that the total number of Lezins and Tabasarns for all these formations except the Kizlar district “exceeds” their total for all of Daghestan while the number of Azerbaijanis in that district should exceed 7,000, “an extremely doubtful” figure.

            Happily for researchers but not so for compilers, nationality data are arrayed in other tables as well. In one, which lists numbers for 13 of the 14 “so-called titular peoples of Daghestan (except the Tats),” the Lezgins number 387,746, 2506 more than the figure released earlier, the Tabasarans 121,809, 2961 more, and the Azerbaijanis 125,452, 5467 fewer.

            What makes these shifts so striking is that they are the only variation from the previously released data and that the total number of the three groups has remained unchanged at 635,007. This does not appear to be an “arithmetic” error, not only because it involves three peoples rather than two but also because it is difficult to believe that the change reflects the lack of data for a single district.

            Rectifying or at least explaining how and why these shifts happened is critical for statistical and political reasons both inside Daghestan and in the Russian Federation as a whole. That is because errors in the data are multiplied whenever they are used together with other data and because government assistance and power flow on the basis of the size of an ethnic group.

            Futulayev gives as an example the way in which republic officials have calculated the growth of various nationalities between the 2002 and 2010 census  According to the data Dagestanstat released earlier, the number of Azerbaijanis in Daghestan increased by 17.3 percent during the intercensal period, the number of Lezgins by 14.4 percent, and the number of Tabasarans 7.9 percent, despite the fact that the last had the highest recorded birthrate.

            Using the new data, these increases were 12.4 percent, 15.2 percent and 10.6 percent, figures that are more plausible but that have the effect of calling into question both the reliability of the data released in June 2012 and the accuracy of the new data which appear to have been manipulated to get a desired result.

            “Whatever were the causes of the distortion of the data of the census on the nationality composition of the population of the Republic of Daghestan,” Futulayev concludes, “clarity must be introduce into this issue” because so much depends upon having accurate numbers in which people can have confidence.



No comments:

Post a Comment