Social Background of the Military

(1)

Social, Health, and Communication Studies Journal

Conflict and the Social Body, Vol. 2(1), December 2015 MacEwan University, Canada

National University of Kyiv-Mohyla Academy, Ukraine Ternopil State Medical University, Ukraine

Article

Social Background of the Military

Conflict in Ukraine: Regional cleavages and geopolitical orientations

Dr. Oleh Ivanov

National University of Kyiv-Mohyla Academy, Ukraine

Abstract

Military conflict in South-Eastern Ukraine is an example of “hybrid warfare” in which “war for minds” is one of the key elements. Differences of Ukrainian people geopolitical orientations often are explained with respect to the region of residence, ethnic identity and native language. Previous research on the problem has three types of limitations: scale, dependent variable (orientations), and regional structures. This study aims to challenge the “on surface” view of place of residence as the key predictor of geopolitical attitudes in Ukraine. The author used excessive amount of data (2005-2015 period surveys with 378,733 cases total sample) and several combinations of dependent variables to test the effects of attitude types, changing political situation and social environment in the regions. The results of regression modeling shows that regional structure is more powerful in explaining general attitude to Russia than integration intentions.

Regional differences proved to be partially explained with the level of lingual- ethnic heterogeneity. In general it seems that language environment has more influence than ethnic. Religious identification also remains very powerful and

(2)

53 Social Background of the Military Conflict in Ukraine

significant. It is claimed that cultural domain of social environment is the key to explain regional structure of geopolitical orientations in Ukraine.

Keywords: geopolitical orientations, regional cleavages, Ukraine-Russia conflict, lingual-ethnic heterogeneity, ethnic identity

_____________________________________

Acknowledgements

I would like to express my deepest gratitude to Dr. Valeriy Khmelko, professor of the National University of Kyiv-Mohyla Academy (NaUKMA), President of Kyiv International Institute of Sociology (KIIS) whose invaluable advice was one of the key driving forces of this research. The data used here was given by him and collected by research fellows and other employees of Kyiv International Institute of Sociology.

_____________________________________

Introduction

Since the USSR collapse, the Black Sea region, in which Ukraine is one of the central powers, has become in many ways important for European security. First, the main energy corridors—gas and oil pipes—are controlled by the states of the region. Second, there were always several “frozen” conflicts in the region which are sources not only of warfare danger (most frequently — Caucasus), but also of crime and humanitarian problems:

smuggling, drugs, weapon and human trafficking. Third, ethnic tensions in the region cause spreading of Islamic fundamentalism, terrorism and other forms of extremism not only throughout Black Sea countries, but at the global level. The fourth factor is the controversial geopolitical position of Russia, which has not only economic and political, but also territorial ambitions. Fifth, the region is essentially trans-boundary: several blocks and military organizations (NATO and CSTO¹ among the main) have their troops ready for ground and sea actions. Therefore, almost all main world powers (including USA, European Union, and Russia) have certain levels of engagement in the regional processes.

1 The Collective Security Treaty Organization. Russia-centered military block.

(3)

All these factors had made the military conflict in Ukraine highly possible, however not expected by many analysts in the near future (Larrabee, 2015, p. 41). The view of Europe as a strategically stable continent has proved to be too optimistic. As a result of the Euromaidan revolution in Ukraine, president Yanukovych was overthrown and sought protection in Russia. The change of Ukrainian authorities extremely reduced the chances of Kremlin to gain control over it via political means. This crisis became a trigger for Russia to act more openly in order to realize its geopolitical aspirations. During March and April 2014 in the Eastern and Southern regions, several propaganda campaigns were organized to separate these regions from Ukraine (Ofcom., 2014; Richter, 2015). After informational preparation, military Special Forces took their turn. They succeeded in rapid annexation of Crimea, but failed in the Donbas region, which resulted in protracted and a slaughterous military crisis.

The conflict in South-East Ukraine represents a new form of “hybrid warfare” — combination of open military actions with undercover special operations, organization and support of separatists, pro-Russian paramilitary groups, combined with an aggressive use of propaganda and disinformation carefully calculated to avoid crossing established thresholds for military response (Hoffman, 2007, p. 18; Larrabee, 2015, p. 22; Bachmann, 2015). The choice of Crimea, Donetsk, and Luhansk regions by Putin’s strategists was not accidental. This part of Ukraine was always more pro-Russian oriented than the West- Northern region. During the Euromaidan unrest in Kyiv, this sentiment became even stronger. Still, the critical point of emphasis is that neither of these regions’ public support of Ukraine’s incorporation into the Russian state was above 50% of the adult population (Kyiv International Institute of Sociology, 2014). Therefore without Russian support these sentiments would have never turned into actions which put Ukrainian sovereignty into question. The key research question of this article is as follows: which social characteristics of South-East Ukraine population made it the most suitable target of Russian “war for minds”?

Previous Research

Taking into account the importance of Ukraine in the European balance of powers, mass political attitudes of its population have been among key objects of post-Soviet studies (see for instance Kulyk, 2009; Katchanovski, 2008; Lane, 2008; Munro, 2007;

Shulman, 2001, 2004, 2005, 2006; Barrington and Herron, 2004; Barrington, Herron, and Silver, 2003; Barrington, 1997, 2001, 2002a, 2002b, 2002c; Kubicek, 2000; Holdar, 1995;

Arel, 1992). Among different social variables taken as explanatory for Ukrainian citizens

(4)

geopolitical differences three proved to be the most influential: region of residence, ethnic identity and native language (Barrington, Faranda, 2009; Smith, 1997; Khmelko, 1998).

The last census of 2001 showed that the dominant (77.8%) nationality in Ukraine is Ukrainian, while 67.5% of those individuals polled named Ukrainian as their native language (State Statistics Committee of Ukraine, 2004). However, more careful question wording and use of question blocks for lingual-ethnic identification in sociological polls of KIIS showed that the population is much more heterogeneous (Khmelko, 2004). KIIS polls showed that Russian speakers are more numerous than Ukrainian and that their proportion is rather stable (Khmelko & Oksamytna, 2008, p. 2).

At the same time, lingual-ethnic heterogeneity highly connected with regional structure: the percentage of Russophones grows from the West through to the Center and from South to East. In turn, regional structure to the most extent defines the distribution of geopolitical orientations and as a result — voting preferences (Khmelko & Oksamytna, 2008, p. 3). The latter tendency became prominent originally during the 1994 presidential elections when leading candidates started to manipulate through pro-/anti-Russian sentiments (Birch, 1995) — political technology has been used in each campaign for more than a decade already (Khmelko, 2006; Frye, 2015).

While region of residence is the most “on surface” factor of foreign policy orientations in Ukraine, there is no consensus among the scholars working in this area of studies. The discussion core is of world-old dilemma: chicken or the egg, namely whether lingual-ethnic heterogeneity defines regional differences in geopolitical orientations or regional structure (which comprise historical background and predominant social environment in each region) defines lingual-ethnic heterogeneity and geopolitical orientations in turn.

All previous studies have their limitations which we can summarize in three blocks:

(1) limitations of scale; (2) limitations of dependent variable (geopolitical orientations) type and (3) limitations of regional structures. If we focus on one of the most statistically rigorous examples — Barrington and Faranda (2009) — these limitations can be specified as follows:

(1) It is done on the basis of only one survey collected in July 2005 by the Kiev International Institute of Sociology. There may be some effects of that time political situation, especially after the Orange revolution which sharpened regional divisions through pro-West and pro-Russian orientation of two main candidates of the 2004 presidential elections in Ukraine.

(5)

(2) The authors used a specially constructed pro-Russian sentiment index variable, which takes into account only the general attitudes to Russia: favorable or unfavorable view of Russia, consideration of its role in world politics, positive or negative influence on Ukraine. Still, there are more definitive attitudes. For instance, the willingness to unite Ukraine and Russia into one state or, at least, eliminate customs borders and visa control. Other orientations of these type include readiness to vote particularly on referendums about NATO, CSTO and other West- or Russia-centered interstate associations. Such orientations are more geopolitical per se than merely sentiment to Russia, and thus potentially more relevant in explaining the support of changing the geopolitical map of the region.

(3) Regions used as inseparable independent variable. The only use of region population characteristics is the interaction of region and lingual-ethnical affiliation of each respondent (individual level of analysis). Still, there may be the influence of social environment. Each region has its dominant ethnic, lingual, and other population categories whose orientations may suppress more marginal categories. Thus, the level of these categories’ heterogeneity in each region, which is rather a macro than an individual indicator, may define the level of geopolitical orientations’ uniformity.

To test these hypotheses, our analysis consists of two parts: first, we replicated the Barrington and Faranda (2009) analysis on different datasets with the use of different types of geopolitical orientations measurement as the dependent variables. Second, we included an explanatory group of variables which indicated the level of lingual-ethnic heterogeneity of the macroregions.

Data Structure

The first test was performed with the use of joint 2005-2015 years dataset. The total sample combines almost 400,000 cases. The set of dependent and independent variables in this array are slightly different than those used by Barrington and Faranda (2009). The religion affiliation is coded as pro-Russian (Orthodox – Moscow Patriarchy) believers, pro-Ukrainian (Orthodox – Kyiv Patriarchy and the Ukrainian Autocephalous Orthodox Church combined), pro-Western (Uniate and Roman Catholic Church combined), other believers and non-believers. The level of education is coded as incomplete secondary education, complete secondary education, special secondary education, and incomplete and complete higher education combined (Barrington and Faranda separate these two

(6)

categories). The locality size is divided only between rural and urban instead of a 6-grade scale. Such coding appears to be more explanatory, as Barrington and Faranda (2009, p.

241) themselves assume only a two-sided contrast: “Ethnic Russians are found more often in larger cities, while rural areas tend to be predominantly made up of ethnic Ukrainians.”

Personal economic standing variable is absent in our joint dataset. However, Barrington and Faranda did not find any significant influence of this variable in all regression models they calculated.

Separate attention should be devoted to native language estimation as it is one of the key explanatory variables. Barrington and Faranda (2009) used one simple question

“Which language do you primarily speak at home?” (Ukrainian, Russian, Both Ukrainian and Russian, Other) (p. 237). Arel and Khmelko (1996) argued that language behavior (the language people actually speak) was a far better indicator than language identity, which was reported by respondents. In this study, we used two-step questioning, developed by Khmelko (2004, p. 5). The interviewer evaluates the language in which the respondent answers to the greeting of “Good Day” pronounced in a way that makes it unclear whether he spoke Ukrainian or Russian and then asks whether he or she would prefer to continue the interview in Ukrainian or Russian. If the answer is “it does not matter,” it is asked which of the two languages the respondent speaks the most. The resulting combined variable consists of six values: Ukrainian; Russian; speaks Ukrainian more; doesn’t matter, speaks both, answers in Ukrainian; speaks Russian more; doesn’t matter, speaks both, answers in Russian. For language effects modeling, it is recoded as “Ukrainian” + “speaks Ukrainian more” — Ukrainian, “Russian” + “speaks Russian more” — Russian. Other respondents were considered as bilingual.

Besides the joint dataset, we also used poll results of April 2008. It consists of 2037 cases and represents public opinion of the later period of the Yushchenko presidency.

Independent variables of this dataset we used for analysis are totally comparable with Barrington’s and Faranda’s (2009) set. (i.e. questions and answers wording are identical).

We calculated separate regression models for two independent variables in this dataset:

(1) Answers to the question “What is your general attitude to Russia now?” (very good, generally good, generally bad, very bad, hard to say). This question is the closest to Barrington’s “general attitude” index.

(7)

(2) Answers to the question “Which relations with Russia do you prefer?” (The same as with other nations – with closed boundaries, visa and customs control;

Ukraine and Russia must be independent but friendly states; Ukraine and Russia must unite into one state.). The question scale was developed by Valeriy Khmelko (2014, p. 4).

The scales of these variables were recoded with the use of Barrington’s approach to constructing the attitude scale, thus the values changed from -10 (pro-Western extreme) to +10 (pro-Russian extreme).

The third dataset included poll results of March 2010. It consisted of 1226 cases and represented the Yanukovych presidency. Independent and dependent variables in this dataset were identical to April 2008 poll.

The last dataset is April 2014 poll, which was conducted after Yanukovych resignation and just as Russian aggression began. It consists of 2022 cases. The predictors set is identical to the previous two polls. We used four types of independent variables in our analysis:

(1) General attitude to Russia—index variable based on four questions—“What is your general attitude to Russia now?”, “What is your general attitude to Russians (Russia residents) now?”, “What is your general attitude to Russia authorities now?”—the scale of answers to these three is “very good, generally good, generally bad, very bad, hard to say”, “Russian society represents essentially different values than Ukraine” (completely agree, partially agree, neither agree nor disagree, partially disagree, completely disagree).

(2) Answers to question “Which relations with Russia do you prefer?”—the same as in previous datasets.

(3) Preferable foreign policy direction—index variable based on 2 statements evaluation—“Ukraine should tighten the links with Europe, even if it worsen its relations with Russia”, “Ukraine should tighten the links with Europe, even if it worsen its relations with Russia”. The answers scale includes 5-components:

completely agree, partially agree, neither agree nor disagree, partially disagree, completely disagree.

(8)

(4) What would you choose on the referendum: Ukraine joining to Customs Union with Russia, Belarus and Kazakhstan or European Union? The question and answers set is identical to the same question in joint dataset.

The level of lingual-ethnic heterogeneity of the macroregions was determined with the set of 9 aggregation variables — percentages of lingual-ethnic groups in each region identified in the polls results. The ethnic identity was revealed with the use of question

“How do you identify yourself by nationality?” (Ukrainian, Russian, the list of “other”

nationalities is provided, but coded as one value). The groups that were considered included the following: Ukrainians—those who identify themselves only with Ukrainian ethnos, Russians, Ukrainian language speakers, mixed Ukraine-Russian language² speakers, Russian language speakers, Ukrainian language speakers and simultaneously Ukrainians, mixed Ukraine-Russian language speakers and simultaneously Ukrainians, Russian language speakers and simultaneously Ukrainians, Russian language speakers and simultaneously Russians.

Results

In total, we calculated 48 ordinary least squares models for the first part of our analysis and 49 models calculated for the second part. All dummy variables excluded as comparative in the first part are the same or analogous to the Barrington and Faranda (2009) approach. In the second part, the set of dummy variables from the first part were combined with the scale variables of lingual-ethnic heterogeneity. Presented below are the key parts of the regression coefficients tables, which are important for our analysis. The whole calculations output, as well as SPSS syntax files of all variables coding and analysis are available from the author, upon request.

The following tables 1-2 include the coefficient estimates and the standard error of each of these estimates. Coefficient estimates with a superscript of “***” represent coefficients significant at the p ≤ .01 level, superscript of “**” represent p ≤ .05 level of significance.

2 Which is often referred to as “surzhyk” — a range of mixed (macaronic) sociolects of Ukrainian and Russian languages.

(9)

Table 1. OLS Regression Coefficient Estimates. April 2014 Poll. Baseline model.

General attitude to Russia (from - 40 to 40)

Which relations with Russia do you prefer?

(from -10 to 10)

Preferable foreign policy direction (from -20 to 20)

Customs Union with Russia or European Union? (from - 10 to 10) B Std.

Err. B Std.

Err.

Constant

term -22.74*** 2.995 -8.07 0.999 -13.54*** 2.033 -9.56*** 1.333 East region 24.335*** 2.001 5.086 0.669 17.375*** 1.359 11.273*** 0.892 Eastcentral 11.541*** 1.871 2.677 0.626 11.66*** 1.271 7.729*** 0.833 South 12.174*** 1.911 1.987** 0.639 11.434*** 1.298 6.592*** 0.853 Southwest 10.108 1.984 1.527** 0.664 8.3*** 1.347 2.111 0.882 Northcentral 2.626 1.643 1.354** 0.55 4.88*** 1.115 2.66*** 0.731 Westcentral 0.585 1.72 0.733 0.575 5.381*** 1.168 1.604 0.767 Russian

ethnic identity 7.993*** 1.229 2.388*** 0.41 4.786*** 0.835 3.014*** 0.549 Other ethnic

identity

4.694*** 1.728 1.652*** 0.576 2.411 1.173 1.771 0.774 Ukrainian and

Russian language speaker

3.479 2.194 -0.578 0.731 1.983 1.49 0.56 0.983

Russian language speaker

6.91*** 1.106 0.644 0.37 4.207*** 0.751 1.8*** 0.495

Orthodox–

Moscow 8.472*** 2.063 1.699** 0.69 2.624 1.401 2.116** 0.918 Orthodox–

Kyiv 5.827*** 1.887 1.363 0.632 2.152 1.281 1.609 0.839 Orthodox–

Other / Orthodox General

3.437 3.037 2.944** 1.013 -2.949 2.062 -0.106 1.35

Other

believer 8.16*** 1.976 1.565 0.661 1.823 1.342 1.257 0.879 Non-believer 6.382*** 2.396 1.022 0.801 1.8 1.627 1.928 1.067 Male

respondent -1.019 0.713 -0.432 0.238 -0.087 0.484 -0.333 0.318

(10)

Higher

education -0.133 2.171 -1.716 0.724 -2.487 1.474 -1.104 0.966 Incomplete

higher education

-3.21 2.802 -1.441 0.934 -5.688*** 1.902 -1.442 1.246

Special secondary

0.754 2.115 -1.054 0.705 -0.705 1.436 -0.232 0.941 Complete

secondary 1.632 2.164 -0.684 0.721 -0.635 1.469 0.14 0.963 Incomplete

secondary 1.134 2.475 0.767 0.828 0.648 1.68 0.823 1.106 Age 30-39 3.274*** 1.12 0.485 0.374 1.31 0.76 0.83 0.5 40-49 2.481** 1.161 -0.365 0.387 0.111 0.788 0.104 0.52 50-59 3.716*** 1.143 0.101 0.382 1.446 0.776 0.922 0.51

>60 3.96*** 1.113 0.807 0.371 2.085*** 0.756 1.286*** 0.498 Village 2.93*** 1.225 0.172 0.41 -0.811 0.832 0.696 0.549 Very small

city 1.286 1.704 -0.84 0.568 -3.735*** 1.157 1.308** 0.761 Small city 7.167*** 2.028 1.008 0.676 1.671 1.377 3.377 0.903 Medium-sized

city (20–99K)

0.414 1.25 -0.187 0.418 -1.203 0.849 0.677 0.56 Large city

(100–499K) 4.176*** 1.189 0.948 0.397 0.366 0.807 0.703 0.533 Enough

money for food but not clothes

-1.833 1.31 -0.564 0.437 -1.945 0.89 -0,815 0,587

Money for food and clothes but not expensive things

-1.908 1.393 0.046 0.465 -2.292 0.946 -1,42** 0,623

Money for some expensive things or whatever the family wants

1.439 1.884 0.526 0.628 -2.092 1.279 -0,918 0,84

(11)

Models with the general attitude dependent variable have the highest R²=0.43 in average, which is practically the same that Barrington and Faranda (2009) had. Moreover, almost all coefficients which are significant in their baseline model were also significant in our model with this variable. In second place are the models with the preferable foreign policy direction dependent variable — R² is nearly 0.4. The other models are substantially less explanatory: R²≤0.2.

The following table represents changes in regional structure influence, if we include regional lingual-ethnic heterogeneity. All other predictors for the corresponding datasets are the same.

Table 2. OLS Regression Coefficient Estimates with lingual-ethnic heterogeneity inclusion.

April 2005 poll. Dependent variable: Which relations with Russia do you prefer? (the value changes from -10 to 10)

Crimea -2,946 0,776

Percentage of russians 0,174*** 0,013

April 2008 poll. Dependent variable: What is your general attitude to Russia now? (the value changes from -10 to 10)

Crimea -5,19*** 0,706

April 2014 poll. Dependent variable: index

“General attitude to Russia” (the value changes from -40 to 40) Model 1

Eastcentral region 1,749 1,148

Model 2

South region 1,908 1,282

Westcentral 2,188 1,347

Percentage of Ukrainian language speakers and simultaneously Ukrainians

-0,177*** 0,018

Model 3

Percentage of bilingual-speaking

ethnic Ukrainians 0,252*** 0,056

(12)

Model 4

Percentage of Russian-language

Model 5

Eastcentral region -8,064*** 1,178

South -5,032*** 1,308

Northcentral -7,278*** 1,034

Percentage of Ukrainian language

speakers -0,291*** 0,018

Model 6

South -4,331*** 1,3

Percentage of Russian language

speakers 0,298c 0,019

Dependent variable: Which relations with Russia do you prefer? (the value changes from -10 to 10)

Model 1

Eastcentral region 0,45 0,383

Model 2

South region -0,401 0,427

Northcentral -0,333 0,33

Percentage of Ukrainian-language

ethnic Ukrainians -0,036 0,006

Model 3

Northcentral -0,153 0,326

Model 4

South -1,696*** 0,437

speakers -0,058*** 0,006

Model 5

South -1,556*** 0,434

(13)

speakers 0,059*** 0,006

Dependent variable: Do you prefer Customs Union with Russia or European Union? (the value changes from -10 to 10) Model 1

Percentage of Ukrainian-language ethnic Ukrainians

-0,098*** 0,008

Model 2

Southwest 1,762 0,808

Model 3

South -1,549*** 0,586

speakers -0,127*** 0,008

Model 4

South -1,243** 0,583

speakers 0,13*** 0,008

The key message we should take from these regression tables is that lingual-ethnic heterogeneity is definitely one of the deeper level explanatory variables of the on-surface regional cleavages in geopolitical orientations. Although the effects of lingual-ethnic groups proportions are not very powerful, they are all highly significant and, what is more important, change regional effects estimates. Those effects become insignificant, change their direction (sign), or both. This tendency is especially true for the South-East half of Ukraine. Unlike previous models with only separate effects of regional structure and lingual-ethnic identity, these estimates are explanatory for all types of geopolitical orientations we consider.

Still there are some peculiarities in each model worth considering:

 The effect of a small fraction of Ukrainians in Crimea diminishes the separate effect of living in that region in relation to attitudes on integration. Crimean

(14)

ethnic Ukrainians favor eurointegration. This tendency is noticeable throughout the decade from 2005-2015 (joint 2005-2015 dataset). The smaller measurement points (2005 and 2008 year datasets) show this tendency for Russians, but obviously indicating the opposite direction.

 The effect of the proportion of Ukrainian speaking ethnic Ukrainians in Crimea, Eastcentral, South, Southwest, and Northcentral regions diminishes the separate effect of living in these regions, if we consider the joint dataset. At the same time, the effect of the proportion of Ukrainian speaking ethnic Ukrainians is observed only in East and South regions.

 The Ukrainian language speakers’ proportion makes insignificant the separate effect of living in Crimea, Eastcentral, South, Southwest, and Northcentral regions. The same effect of Russian-speaking percentage is present in Crimea, Eastcentral, and South regions only.

While comparing three periods of Ukrainian political history after 2005, we note the following peculiarities:

 During the Yushchenko and Yanukovych presidencies, the tendencies of lingual- ethnic heterogeneity influence are essentially the same. The exceptional is only the model which takes into account the percentage of Russian-speaking Ukrainians. During the years 2005-2009, the separate place of residence predictor in South region is insignificant, while between 2010 and 2013 — it is lower.

 After 2014, the estimated coefficients of regions should be considered, while taking into account the absence of Crimea in the samples. Therefore, certain transformation of the regional scale of geopolitical orientations can be seen.

Especially noticeable is the Southwest region, where lingual-ethnic groups’

distribution diminishes the residence effect in the majority of models.

Discussion

The significance of the coefficients in the joint dataset may be an answer to Barrington’s and Faranda’s (2009) claim: “Another question for further study involves whether or not the weak interaction effects were the product of large standard errors as much as weak effects” (p. 251). They used a dataset composed of 1200 cases, while part of our dataset with data valid for analysis comprised more than 120,000 cases. The lower

(15)

explanatory power of the models with the integration attitude index may be explained by peculiarity of this attitude when compared to the general sentiment.

This assumption was confirmed with the analysis of single datasets, especially the April 2014 survey, which had the most complete block of questions available for the calculation of the attitude index. This set of predictors provides the best explanation regarding the general attitude towards Russia. Geopolitical orientations about real foreign policy actions towards different unions, especially if this unions mean loose of Ukrainian sovereignty, may be better explained with some other predictors. We suggest as one of these predictors, satisfaction with Ukrainian government policy in different spheres, especially the meeting of Russophones’ cultural needs.

At the same time, combined indexes, which allow the capturing of consistent favor towards one of the geopolitical poles, are better explained. The answers to single questions are more dispersed. This suggests that structural influence should be preferably analyzed with the use of carefully calculated indexes, rather than single variables like voting preferences on the referendum.

Coefficients of living in Crimea and Eastern regions suggest that Russian speakers of Eastern regions are more supportive to a union with Russia than Crimean people.

Barrington’s results for these two regions were quite opposite. The difference may be explained, in our opinion, by three factors: (1) the pro-Ukrainian role of Crimean Tartars, based on the decade surveys turn out to be more salient; (2) absence of Crimea in the survey samples after it was annexed by Russia; (3) radicalization of the Eastern region in the integration attitude dimension after 2005, especially after military conflict started in the region in 2014.

The assumption about the possible explanation of regional differences by social environment, namely lingual-ethnic domination of certain groups, is partially confirmed.

In general, it seems that language environment has more influence than ethnic. Still, there are other variables with values across regions that may be part of the puzzle. One of them is religious identification which remains very powerful and significant. This suggests that the cultural domain of social environment is the key to explaining regional structures of geopolitical orientations in Ukraine. By “cultural domain,” we mean the main humanitarian differences of West-Central and East-South parts of Ukrainian societies:

spoken language, ethnic self-identification, and their combinations, social values, confessional affiliation. Some of these variables go beyond the scope of this study, thus they are to be considered in further research.

(16)

It must be noted that socio-cultural and orientation cleavages between two parts of Ukraine are the prerequisites for the conflict and their social background, but not the triggers and direct causes of the military action. Other political, economic, and subjective (personal influence of certain stakeholders) factors, both internal and external, took their turn in transforming disagreement into separatists movements.

Regional elites of south-eastern Ukraine sought control over other regions of the country, but because of the Euromaidan, they faced the threat of losing influence even in Donbass (Skvorets, 2015). Eastern Ukraine over the past decades had a parallel government system, popularly called “oligarkhat” (“oligarch ruling”). The 2010-2013 period was marked by the attempt of eastern oligarchic elites to expand their control on the territory of the whole country. By using politically motivated appointments (mostly of people from Donbass: Donetsk and Luhansk regions) to senior positions in regional units of the Ministry of Internal Affairs, Security Service of Ukraine, prosecutors, tax, customs and justice officials, they subordinated regional business and political elites, and co-opted them into the political structure of the Party of Regions (Piechal, 2015). The success of this tactic in various regions was different. In western regions, it did not lead to an appreciable success, but on the contrary, has become one of the main reasons of the Euromaidan and its mass support among the residents of the western regions of country.

In Crimea, this tactic led to the actual capture of this autonomous region by oligarchic groups so-called “Macedonians” (people of Makiivka and Donetsk origin) and removal of the local elite from regional executive government institutions (Fisun, 2015). The latter was among the most recent causes of Crimean people growing discontent with Ukrainian central government, ultimately led to popular support of Russian annexation.

This rough Donbass-centered policy of Yanukovych government eventually led to the formation of separatist paramilitary groups in Eastern Ukraine, when Euromaidan clashes in Kyiv became bloody violent. Starting from Kharkiv “Oplot” organization members, their supporters claimed that a paramilitary coup d’etat had happened in Kyiv (Kuzio, 2015; Laruelle, 2015). From the Donbass people’s point of view their representative overthrow looked absolutely unlawful and intolerable. The position of Donbass regional elites was somewhat different: they wanted to use separatist movements to bargain with new Kyiv authorities, claiming to preserve their exclusive status (Golovakha, 2014). After Russia’s intrusion, when the unrest in Donbass transformed into the Ukraine-Russia conflict, oligarchs retrieved to Kyiv, leaving the region under the control of paramilitary groups’ commanders and Russian emissaries (Piechal, 2015).

(17)

Conclusion

As a result of replication of Barrington and Faranda’s approach on the datasets of different periods, scale, dependent single-question variables and combined indexes, we can confirm that on the individual level of analysis “region of residence” is a key explanatory variable of geopolitical orientations. Still, the regional structure is more powerful in explaining the general attitude to Russia than integration intentions. Also, our analysis suggests that the Donbass region is more radical and solid if we take its pro- Russian population fraction, even compared to Crimea.

Using lingual-ethnic heterogeneity, predictors in regression models challenge previously dominated view of residence place as the key and inseparable predictor of geopolitical orientations in Ukraine. It is worth moving in the direction of studying wider cultural heterogeneity of the regions. It is an object of change unlike region boundaries, which are established administratively. Thus these predictors expected to be more valuable for prognosis than the rigid regional structure.

In further research, it may be valuable to include other control variables, like government policy satisfaction, and use more consistent orientation indexes. A qualitatively upper level of results may be achieved, if we move from analyzing individual respondents’ answers to macro-characteristics of the regions (not upper than oblast level).

That is regions must be considered as cases in model calculations, and aggregated indicators (proportion of ethno-lingual groups, level of urbanization etc.) as predictor variables in regression. In that case, we could explain the essence of a regional structure influence, which can only be postulated from individual level regression models.

In this article, we considered internal social prerequisites of the military conflict in Ukrainian South-East: for at least ten years Ukrainian society was, in fact, divided into two almost conflicting parts, mainly on a regional basis. The most pro-Russian sentiments were widespread in Crimea and eastern Ukraine—Donetsk and Luhansk regions. The main reasons of such geopolitical orientations differences are of cultural origin: language, ethnicity, values, and confessional affiliation. However, transformation of this disagreement into military conflict needed the confluence of a number of internal and external factors: Donbass-centered policy of Yanukovych government, Euromaidan, South-Eastern “oligarchs” aspirations, and Russian intervention with the use of hybrid warfare means.

(18)

__________________________________________________

Oleh Ivanov, PhD. is a Senior Lecturer at the Department of Sociology, National University “Kyiv-Mohyla Academy,” and analyst at the Presidential Administration, Ukraine. In 2012, defended “summa cum laude”

PhD thesis “Representation of People’s Foreign Policy Orientations in the Public Communication of Power Elites.”

Scholar interests cover political sociology, sociology of mass communication, content-analytical methodology.

__________________________________________________

References

Arel, D. (1992, November 19–22). Federalism and the Language Factor in Ukraine. Paper Presented at the American Association for the Advancement of Slavic Studies annual meeting, Phoenix, AZ.

Arel, D., Khmelko, V. (1996). The Russian Factor and Territorial Polarization in Ukraine, The Harriman Review, 9 (1–2), 81–91.

Barrington, L. (1997, December). The Geographic Component of Mass Attitudes in Ukraine, Post-Soviet Geography and Economics, 38 (10), 601–614.

Barrington, L. (2001, April–June). Russian-speakers in Ukraine and Kazakhstan:

‘Nationality,’ ‘Population,’ or Neither? Post-Soviet Affairs, 17 (2), 129–158.

Barrington, L. (2002). Examining Rival Theories of Demographic Influences on Political Support: The Power of Regional, Ethnic, and Linguistic Divisions in Ukraine, European Journal of Political Research, 41 (4), 455–491.

Barrington, L. W. (2002). Region, Language, and Nationality: Rethinking Support in Ukraine for Maintaining Distance from Russia. In T. Kuzio & P. D’Anieri (Eds.), Dilemmas of State-Led Nation Building in Ukraine. Westport, CT: Praeger.

Barrington, L. W. (2002, Summer). Views of the ‘Ethnic Other’ in Ukraine, Nationalism and Ethnic Politics, 8 (2), 83–96.

Barrington, L. W., Herron, E. S. (2004, March). One Ukraine or Many? : Regionalism in Ukraine and Its Political Consequences, Nationalities Papers, 32 (1), 53–86.

Barrington, L. W., Herron, E. S., Silver, D. B. (2003, January). The Motherland is Calling: Views of Homeland among Russians in the Near Abroad, World Politics, 55 (2), 290–313. Retrieved from the Web June 24, 2015.

http://www.jstor.org/stable/pdf/25054221.pdf

(19)

Barrington, L., Faranda, R. (2009). Reexamining Region, Ethnicity, and Language in Ukraine, Post-Soviet Affairs, 25 (3), 232–256.

Birch, S. (1995). The Ukrainian parliamentary and presidential elections of 1994. Electoral Studies, 14(1), 93-99.

Frye, T. (2015). What Do Voters in Ukraine Want?: A Survey Experiment on Candidate Ethnicity, Language, and Policy Orientation. Problems of Post-Communism, 62(5), 247- 257.

Holdar, S. (1995, February). Torn between East and West: The Regional Factor in Ukrainian Politics, Post-Soviet Geography, 36 (2), 112–132.

Katchanovski, I. (2008, October-December). The Orange Evolution? The ‘Orange Revolution’ and Political. Post-Soviet Affairs, 24 (4), 351–382.

Khmelko, V. (2004). Ethno-linguistic structure of Ukraine: regional features and tendencies of changes during independence [Ukrainian language article], Naukovi zapysky NaUKMA, 32, 3–15. Retrieved from the Web June 24, 2015.

http://elib.ukma.edu.ua/NZ/NZV32_2004_sociol/01_khmelko_vye.pdf Khmelko, V. (2006). What helps politicians to divide Ukraine? [Ukrainian language

article], Dzerkalo Tyzhnia. Retrieved from the Web October 29, 2015.

http://gazeta.dt.ua/ARCHIVE/cherez_scho_politikam_vdaetsya_rozkolyuvati_ukr ayinu.html

Khmelko, V. (2014). Dynamics of the public support for Ukraine’s sovereignty in the first decades of its state independence (1991–2014 years) [Ukrainian language article], Naukovi zapysky NaUKMA, 161, 3–10. Retrieved from the Web June 24, 2015.

http://www.ekmair.ukma.edu.ua/bitstream/handle/123456789/3191/Khmelko_D ynamika_pidtrymky_naselenniam.pdf

Khmelko, V., Oksamytna, S. (2008, October 21–24). Russian-Ukrainian Biethnical Identity and Lingual-Ethnic Heterogeneity in Ukraine [Russian language paper].

Paper Presented at the Third All-Russia Sociological Congress, Moscow. Retrieved from the Web June 24, 2015. http://www.isras.ru/publications_bank/1225745240.pdf

(20)

Khmelko, V., Wilson, A. L. (1998). Regionalism and Ethnic and Linguistic Cleavages in Ukraine. In T. Kuzio & P. D’Anieri (Eds.), Contemporary Ukraine: dynamics of Post- Soviet transformation. Armonk, N.Y: M.E. Sharpe.

Kubicek, P. (2000, March). Regional Polarisation in Ukraine: Public Opinion, Voting and Legislative Behaviour, Europe-Asia Studies, 52 (2), 295–317. Retrieved from the Web June 24, 2015. http://www.jstor.org/stable/pdfplus/10.2307/153434.pdf

Kulyk, V. (2009, April 23-25). Language Policy in Ukraine: Why People Think the Way They. Paper Presented at the Association for the Study of Nationalities Annual Meeting, New York. Retrieved from the Web June 24, 2015.

https://www.academia.edu/1337259/Language_Policy_in_Ukraine_Why_People_

Think_the_Way_They_Do

Kyiv International Institute of Sociology. How relations between Ukraine and Russia should look like? Public opinion poll results. (2014, March 4). Retrieved from the Web June 24, 2015.

http://www.kiis.com.ua/?lang=eng&cat=reports&id=236&page=1 Lane, D. (2008, November). The Orange Revolution: ‘People’s Revolution’ or

Revolutionary Coup? British Journal of Politics and International Relations, 10 (4), 525–

549.

Larrabee, S. F., Wilson, P. A, & Gordon IV, J. (2015). The Ukrainian Crisis and European Security: Implications for the United States and U.S. Army (67 p.). Santa Monica,

California: RAND Corporation. Retrieved from the Web June 24, 2015.

http://www.rand.org/content/dam/rand/pubs/research_reports/RR900/RR903/

RAND_RR903.pdf

Munro, N. (2007, November–December). Which Way Does Ukraine Face?: Popular Orientations Toward Russia and Western Europe, Problems of Post-Communism, 54 (6), 43–58.

Richter, A. G. (2015). Legal Response to Propaganda Broadcasts Related to Crisis In and Around Ukraine, 2014–2015. International Journal of Communication, 9, 21. Retrieved from the Web October 29, 2015.

http://ijoc.org/index.php/ijoc/article/download/4149/1474

(21)

Ofcom. (2014, November 10). Standards case of RT. Ofcom Broadcast Bulletin, 266.

Retrieved from the Web October 29, 2015.

http://stakeholders.ofcom.org.uk/binaries/enforcement/broadcast- bulletins/obb266/obb266.pdf

Pirie, P. (1996). National identity and politics in Southern and Eastern Ukraine, Europe- Asia Studies, 48 (7), 1079–1104. Retrieved from the Web June 24, 2015.

http://www.jstor.org/stable/pdfplus/10.2307/153099.pdf

Shulman, S. (2001, Summer–Winter). Region, Identity, and Political Authority in Ukraine, Journal of Ukrainian Studies, 26 (1-2), 175–196.

Shulman, S. (2004, January). The Contours of Civic and Ethnic National Identification in Ukraine, Europe-Asia Studies, 56 (1), 34–56. Retrieved from the Web June 24, 2015.

Shulman, S. (2005, Spring). National Identity and Public Support for Political and Economic Reform in Ukraine, Slavic Review, 64 (1), 59–87. Retrieved from the Web June 24, 2015. http://www.jstor.org/stable/pdfplus/10.2307/3650067.pdf Shulman, S. (2006, June). Cultural Comparisons and Their Consequences for

Nationhood in Ukraine, Communist and Post-Communist Studies, 39 (2), 247–263.

Bachmann, S. D. O. V., & Gunneriusson, H. (2015). Russia’s Hybrid Warfare in the East: Using the Information Sphere as Integral to Hybrid Warfare. Georgetown Journal of International Affairs-International Engagement on Cyber V: Securing Critical Infrastructure.

McCuen, J. J., colonel (2008). Hybrid Wars. Military Review, March-April 2008, 107–113.

Skvorets, V. (2015). Internal prerequisites of civilization conflict in Ukraine [Ukrainian language article], Skhid, 39 (2), 102–107.

Fisun O., Krysenko O. (2015). The causes of destabilization in the southern and eastern regions of Ukraine [Russian language article], Kennan Institute Kyiv Office.

Kuzio, T. (2015). Competing Nationalisms, Euromaidan, and the Russian‐Ukrainian Conflict. Studies in Ethnicity and Nationalism, 15(1), 157-169.

Laruelle, M. (2015). The three colors of Novorossiya, or the Russian nationalist mythmaking of the Ukrainian crisis. Post-Soviet Affairs, (ahead-of-print), 1-20.

(22)

Golovakha, Y. (2014) How to stop “war in minds”? [Ukrainian language interview], Holos Ukrainy. Retrieved from the Web October 29, 2015.

http://golosukraine.com/publication/main/statti/35265-yak-pripiniti-vijnu-u- golovah/

Piechal, T. (2015). The War republics in the Donbas one year after the outbreak of the conflict. OSW Commentary, 174. Retrieved from the Web October 29, 2015.

http://aei.pitt.edu/65608/1/commentary_174.pdf

Hoffman, F. (2007). Conflict in the 21st Century: The Rise of Hybrid War (70 p.). Arlington:

Potomac Institute for Policy Studies. Retrieved from the Web October 29, 2015.

http://www.potomacinstitute.org/images/stories/publications/potomac_hybridwa r_0108.pdf

Smith, G., Wilson A. (1997). Rethinking Russia Post‐Soviet Diaspora: The Potential for Political Mobilisation in Eastern Ukraine and North‐East Estonia, Europe-Asia Studies, 49 (5), 845–864. Retrieved from the Web June 24, 2015.

State Statistics Committee of Ukraine. Population distribution by nationality and native language.

[Ukrainian language web-page] (2004). Retrieved from the Web June 24, 2015.

http://2001.ukrcensus.gov.ua/results/nationality_population/nationality_popul1/s elect_5/?data1=1&box=5.1W&rz=1_1&rz_b=2_1&k_t=00&botton=cens_db