Ara Marjanyan Senior Expert, «Noravank» Foundation; UN National Expert on Energy and Transport in Armenia, Ph.D, Senior Scientific Researcher Currently South Caucasus region is in an ongoing information and psychological warfare. One of its characteristics is the systematic misrepresentation of demographic statistics and wide use of distorted data to carry out psychological warfare. Without exaggeration every day it is possible to read, hear and see in the Azerbaijani press and electronic media, in statements of official, semi-official and non-official representatives, in the TV channels and radio broadcasts such phrases as “Armenia is getting emptied”, “9-million-strong Azerbaijan will not tolerate a loss of its territories”, “soon, very soon Azerbaijan will emerge as even a more powerful, 10-million-strong nation”, and finally, that “it is time to establish a Pan-Turkic state with population of 110 million, comprised of 10-million-strong Azerbaijan, 30 million Azerbaijani population of northern Iran and 70-million-strong Turkey” and that “not only the region, but also the whole world will have to reckon with such a giant”. Demographics of the Official Baku In our previous studies [1-3] we analyze the falsifications of official Baku regarding the statistical data on demographics of the Azerbaijani Republic (AR). Let us briefly recap the main conclusions of these studies. Official sources, such as the State Statistical Committee of the Azerbaijani Republic (ASSC), constantly distort the patterns of migration flows and mortality/fertility rates. The official population number exceeds the actual number by 2.0-2.2 million. In 2011 it comprised 6.9-7.5 million, in contrast to the official number of 9.1 million. ASSC data on net migration (i.e. emigration minus immigration) underestimate the real numbers by dozens of times. The number of migrants from AR amounted to about 2.0-2.2 million in 2010-11. Since 1996-97 the official numbers on annual growth rate of population were overestimated by 3-4 times in late 90s - early 2000s, and at least 1.5-2.0 times today. Data on the AR population distribution by sex and age are also blatantly garbled. The number of the female population was artificially increased by 0.5 million, to make the claims of “10-million–strong Azerbaijan” “biologically plausible” to. The April 2009 population census did not enhance the situation with the statistical data in AR. Population of Azerbaijan in 2100 according to the UN DESA Population Division Unfortunately, meddling with demographic statistics is not restricted to the official Baku or Azerbaijani media. Manipulations of official Baku related to demographic statistics in some cases are replicated by certain international organizations. In case of the UN DESA Population Division (UN PD), these manipulations are in a way “validated” in the UN Demographic Yearbooks and widely distributed through electronic and other mass media. Indeed, during the last 20 years the UN PD presents demographic data for almost all countries of the world in its well-known online database1. The UN PD database also allows modeling the population dynamics for countries till the year 2100 in three population growth scenarios: high, medium and low. These dynamics can be not only modeled, but also visualized for a given country or group of countries2. So what pattern for the three South Caucasus countries will appear before the eyes of a curious user of this database? The result for the period of 1950-2100 is shown in Figure 1. As seen, according to the UN PD experts, in case of the high growth scenario by the end of the 21st century the AR population will comprise 19.4 million, whereas that for Armenia and Georgia will amount to 4.0 and 4.1 million, respectively (the curves for Georgia are not shown in Figure 15 for simplicity. The details of modeling results for all considered scenarios are presented in Annex 3). Hence, in this scenario by 2100 the population of Azerbaijan will exceed that of Armenia and Georgia in aggregate by more than twice. The situation is even more “dismal” for the low growth scenario. In this case the AR population (7 million) will be more than 3 times larger than that of Armenia and Georgia together (1.2 million each). Sources: ` http://esa.un.org/unpd/wpp/unpp/panel_population.htm and our own estimates. For an unprepared user of the mentioned database, who is unfamiliar with the long history of systematic manipulations by official Baku and is uninitiated in “peculiarities” of reporting statistical data on AR population in the UN PD publications, this pattern may be shocking. It has to be noted that the UN PD database is a popular source for numerous publications around the world, including Armenia. It is almost the only international source for the publications in Azerbaijani media. And this pattern remains widely accessible at the UN PD website, without any comments or caveats. Yet comments here would have been more than relevant, because the presented information contains some unjustified assumptions, distortions and falsifications. First of all, the high growth scenario for AR population 2015-2100 is generally linear. Moreover, for one and a half century period of 1950-2100 it can be approximated by a simple linear equation with accuracy of 99.78% (see the green line on Figure 1). This is remarkably similar to the behavior of the official data presented by ASSC for 1950-2011 (see [3], Fig. 7). It appears, there is every reason to argue that the UN PD experts simply replicate the unrealistic and primitive linear model of population growth used by ASSC “specialists” for many decades (1950-2011), thus unblinkingly applying it for the whole period of 2015-2100. Besides, the UN PD experts use only official ASSC data on population, without any examination or comments. However, the comments would have been more than appropriate, at least with regard to the blatant distortion of the AR population numbers and distribution by age and sex, which are typical to both ASSC data and the UN PD Demographic Year-Books (see [3], Table 2 above and below). Finally, the actual population number curve for Armenia (as well as for Georgia) in the period prior to 1995 has a characteristic breakpoint at around 1990-95 (see Figure 1, marked with an ellipse). Indeed, the breakup of the USSR, collapse of the socio-economic system, onset of regional conflicts and intensive processes of migration inside and outside of the region, at the time led to significant changes in the demographic situation of all South Caucasus republics. Then why all of this passed unnoticed in case of the Azerbaijani Republic’s population? The answer that one might get from the UN PD experts is “because these are the official data, received from the country’s official statistical services.” Incidentally, the UN PD experts have diligently accounted for the retroactive recalculation of Armenia’s population number performed by the National Statistical Service of the Republic of Armenia (RoA) for the period between 2002 and 1990 (see [3], Fig. 3) resulting from the 2001 RoA census. Therefore, it can be stated that in their same database the UN PD experts on one hand, in case of Armenia, quite meticulously take into account demographic statistical recalculation and corrective improvement based on the 2001 RoA census results, and on the other hand, in case of Azerbaijan, they cool-headedly repeat absurd official numbers of the country’s 1999 and 2009 censuses. All of this is done without any comments and assessments regarding the reliability and validity of the presented data. Yet the most ironic aspect in the UN PD population models is the modeling/forecasting of migration flows for the three South Caucasus republics, based on net migration. As a reminder, a negative value of net migration means prevalence of emigration (population outflow), whereas positive value of net migration indicates prevalence of immigration (population inflow). The analyses of related portion of UN PD database pose several alarming questions. So how come for the period of 1990-95 net migration for Armenia and Georgia was estimated at about -110,000 annually each, while for Azerbaijan it was -23,000? How come for the period of 1995-2000 net migration for Armenia was estimated at -45,000 (and -78,000 for Georgia), whereas for AR it was -26,000? An apparent answer from UN PD experts in case of Azerbaijan would be: “because that is the official statistics of Azerbaijan”. However, the official statistics of Armenia for those years is different. Then why the UN PD experts do not base their estimates on official statistics in case of RoA, as they do it in case of AR? After all, are the UN PD experts unaware that according to a number of reputable international organizations3 by early 2008 there were 1.5 million emigrants from Azerbaijan residing abroad? Do they not know that in 2010 the emigrant stock of the AR remitted about US$1.5 billion from abroad to Azerbaijan (see [3], Tables 4 and 5) through the international banking system? And this is the amount of remittances adjusted for purchasing power parity. The actual amount (in current dollars) is about 1.5 times larger. And this concerns only those Azerbaijani emigrants who made remittances to Azerbaijan in a given year using the services of the international banking system. Certainly there were also emigrants who transferred money to the country using other channels, or did not make any remittances at all. The actual number of emigrants from Azerbaijan is a lot higher and according to some reliable estimates, by 2010 it constituted about 2.0-2.2 million. A sizeable portion of this emigrant stock was formed in 1995-2005 (see [3], Ch. 2). How could this emigrant stock appear under such absurdly low emigration numbers used by UN PD? According to the UN PD, net migration of Azerbaijan for the period of 2000-2010 in average comprised about +11,000 annually (+110,000 over the period of 10 years in 2001-2010). It would be interesting to know why. This time even the standard answer “because that is the official Baku statistics” would not help the UN PD experts. That is because even the data from official Baku are different in this case. Let us first recall that as per ASSC approach for calculating net migration, only those arriving or departing for permanent residence are accounted for (see [3], Ch. 2 or the ASSC website). We shall not reprove the UN PD experts for turning a blind eye on such an odd and unprecedented approach4. We shall rather invite their attention to the fact, that even with such approach official Baku estimated the net migration for 2001-2010 at about -14 000, or -1.4 thousand annually in average (see [3], Fig. 4). Then why the UN PD experts in their models deepen the already absurd situation and bring the net migration number for 2001-2010 up to +110 thousand, surpassing even official Baku in this aspect? If all of this is ironic, then the way in which UN PD database forecasts net migration in 2010-2100 for South Caucasus countries is simply ridiculous. Why net migration for Azerbaijani Republic equals to zero for the whole period of 2015-2100? Why emigration from Georgia stops in 2055 and why it continues for Armenia till 2100? These are questions, which one may presume can be answered only by official Baku, but not the UN PD experts who act as obedient stooges of latter. Population of Azerbaijan in 2100: Corrected Picture In summary, it seems necessary to raise the following questions for the authors of the above mentioned online database of the UN PD. First, if this database simply repeats the official statistical data from countries, then why is it needed in the first place? Such data are accessible at national statistical agencies and a source that simply repeats them appears redundant5. Even without that official Baku already floods the internet with falsified demographic statistics, and intensively uses that in the information and psychological warfare. Second, if the mentioned UN PD database is intended to do more, for example, to model/forecast the population for long future periods, then why the UN PD experts seem so unconcerned about the reliability of their models? Surely it is impossible to perform adequate modeling if the input data are known to be garbled. For a more or less knowledgeable specialist in computer simulation and forecasting it is a given, that estimating of reliabilityand validity of the input data comes first. If it is impossible to improve that by own means or desire, then the uncertainties in the modeling must be estimated and taken into account first, and afterwards the results should categorized according to their reliability. Third, do the UN PD experts realize that by ignoring the issue of reliability/validity and using only the estimates of the national statistical services as input data, they so to speak “validate” and amplify both positive and negative aspects and trends of those? Such approach fails to amend the inferior and make the good better. Figuratively speaking it makes the strong stronger, the weak weaker, the frank franker, and the cunning and deceitful more cunning and deceitful. Is this really the mission of the UN? Since there is little hope that the importance of these matters will be realized by the UN PD experts in a more or less reasonably short period, we undertook here to correct the situation with the case of Azerbaijani Republic. For this purpose, first of all the above reviewed manipulations of official Baku related to the actual migration in the AR prior to 2012, actual population growth rate and actual sex and age distribution of the population were accounted for. Also, the bizarre assumptions of the UN PD regarding migration flows of South Caucasus republics for the period of 2012-2100 were revised. The corrected results are presented in Figure 1 (see the red curves). They suggest that by 2020 the actual de facto population of AR will comprise about 7.6-8.03 million people. By 2100, the AR population will constitute 12 million in “high” growth scenario (denoted as “AR, corr. high” in Figure 1), 7 million under the “medium” growth scenario and 3.5 million under the “low” growth (“AR, corr. medium” and (“AR, corr. low” in Figure 1). The corrections lead also to a behavior of the AR population curve for 1990-2000 which is characteristic to those of almost all post-Soviet countries. It is obvious that for long-term forecasts of a country’s population applying primitive, linear models is unacceptable. This is what has been learned from the world experience of complex system simulation since first attempts of the Club of Rome. The internal socio-economic tensions, external impacts, various non-linear effects, so called “wild cards” and “phase transitions” have to be taken into consideration. Recently we witnessed a number of such “phase transitions”. For instance, today it is ironic to see the sleek and nice population curves for Sudan in the UN PD database that reach 2100. In reality, Sudan has undergone a “phase transition” and no longer exists. It has broken up in two parts, which are immersed in deep conflict and “chaos”. The same can be told about some other countries. And most likely, more such examples will be seen in the near future. However, it appears that such issues are completely beyond the reach of thought of the experts who created the mentioned UN PD online database. As a final remark, our corrections of the UN PD forecasts of the AR population till 2100 are not aimed at shadowing the issue of current demographic situation in Armenia. On the contrary, we believe this is a pressing and urgent issue, but it requires deep, somber and comprehensive study, mostly focused on developing measures to overcome the existing negative trends. 1 http://esa.un.org/unpd/wpp/unpp/panel_population.htm 2We would like to caution the visitors of the UN PD electronic library to be wary of the good-looking charts and high-sounding names (such as double logarithmic fertility curve, Bayesian Hierarchical Models, etc.). As we shall see later, all of this is rather senseless in case of the AR. As the saying popular among the modeling professionals goes, “garbage in, garbage out”, no matter how complex and sophisticated your model is. 3IOM, WB DPG, etc. 4Though, sooner or later, somebody neutral will hopefully do that. 5Even more so, as they are repeated in the UN Demographic Yearbooks. References and Literature ????????? ?., ???????????? ?????????????? ??????????? 2009?. ??????????? ???????????? ?????, 19.12.2011.http://noravank.am/arm/articles/detail.php?ELEMENT_ID=6186&sphrase_id=14684 ????????? ?., ???????????? ?????????????? ??????????? ??? ????? ?? ??????? ????? ?????, 22.02.2012.http://noravank.am/arm/articles/detail.php?ELEMENT_ID=6319&sphrase_id=14684 ????????? ?., ???????????? ?????????????? ???????? ????????????????, ??????, «????????» ???, 2012?.,http://noravank.am/upload/pdf/Ara%20Marjanyan%20book.pdf “Globus” analytical journal, # 7, 2012

---