It is well known that Russian President Vladimir Putin’s actions in Ukraine, annexing Crimea in March 2014, and sponsoring rebels thereafter in eastern Ukraine, generated high levels of support for him in Russia in 2014. But how popular is Putin in contested Ukraine, namely in Crimea and the southeast swath of territory he described in a talk show in April 2014 as “Novorossiya”? The nearest modern equivalent to this contested historical region is a semi-circle of territory comprising eight of Ukraine’s 24 oblasts in the south and east of the country: Odessa, Mykoliav, Kherson, Dnipropetrovsk, Zaporizhia, Kharkiv and the two oblasts racked by fighting since April, Donetsk and Luhansk.

Conventional wisdom is that support for Putin is related to the percentage of ethnic Russians in a region and the percentage of the population who declared that their native language is Russian. It is common to find choropleth maps of both language and national distributions by oblast from the last Ukraine census in 2001 in mass media features on Ukraine. These maps show how concentrations of Russians and Russians speakers are highest in Crimea, followed by Donetsk and Luhansk and then other oblasts in so-called “Novorossiya”. But is the popularity of leading political figures like Putin or the current Ukrainian president Petro Poroshenko really determined by ethnicity and language use in Ukraine?

With support from the Political Science program of the U.S. National Science Foundation, we organized surveys in southeast Ukraine and Russian-controlled Crimea in December 2014. The door-step surveys are extensive with more than 140 questions on geopolitical developments, perceptions of the current conflict, preferences for political arrangements, attitudes towards various groups, ratings of political figures, self-declared identities and reconciliation possibilities. The Kyiv International Institute of Sociology administered our survey to 2,000 persons in six oblasts in southeast Ukraine. (We had to exclude the war zones of Luhansk and Donetsk because face-to-face surveys with a rigorous sampling design are currently impossible there). At the same time in late December, the Levada Center in Moscow administered a similar survey to 752 respondents in Crimea using the same interview procedures.  Over 120 of the 140 survey questions are matched thus facilitating comparative analysis of responses by nationality and language groups in the two settings. Tatars (10.2 percent of the population in 2001) as well the ethnic Russian majority and the large ethnic Ukrainian minority were sampled in the Crimean survey.

To elicit an instantaneous measure of political popularity, we asked respondents in the survey to review four black and white head-shot photographs without identification. We theorized that an image rather than a name would more directly access the affective dispositions crucial in conflict contexts. Respondents were simply asked how they felt about the ‘man in the photo’ on a five point like/dislike scale. (‘Difficult to answer,’ ‘do not know who the man is’ and ‘refuse’ were also options, though less than 5 percent chose these answers). In both surveys the first image was Vladimir Lenin, the third Barack Obama, and the fourth Vladimir Putin. In southeast Ukraine, we used an image of Poroshenko, while in Crimea, we used an image of the Russian-backed leader Sergei Aksyonov.

Allowing respondents to self-identify their nationality, the survey also asks about the language, or languages, that the respondent primarily spoke at home. This provided us with a measure of actual language use as distinct from the census category ‘native language’. Though 42 percent reported that Ukrainian is their native tongue, just over half that ratio use it as the primary home language.  In southeast Ukraine, we were thus able to isolate four predominant identity categories: (i) ethnic Ukrainians speaking Ukrainian (22.6 percent of our sample), (ii) ethnic Ukrainians speaking both Ukrainian and Russian (17.4 percent), (iii) ethnic Ukrainians speaking Russian (40.7 percent) and (iv) ethnic Russians speaking Russian (11.6 percent). The Crimea survey featured three ethnic groups: ethnic Russians (63.5 percent), ethnic Ukrainians (20.9 percent) and Tatars (8.5 percent). Almost 85 percent of this sample, however, spoke Russian at home, with 7.2 percent indicating they spoke both Russian and Ukrainian and 7 percent Tatar. In the figures, we display Crimean results using just ethnicity.

Our political popularity survey scores are shown with proportions for each of the possible respondents for each photo.  Generally, about one-third of respondents gave the politicians neutral scores though a half of respondents in southeast Ukraine gave that value to Obama and a similar high ratio gave to Lenin in Crimea.  Putin clearly generated more diversity of opinion with high negative scores in the six Ukrainian oblasts and high positive ratings in Crimea.

The last series of graphs display a composite score that we call ‘likeability’ generated from the weighting of values for each response (+2 to -2) multiplied by the proportions of the respective national-languages that chose that option.  The consistency of the scores of the language-national groups in southeast Ukraine stand in sharp contrast to the variations by politician and nationalities in Crimea.

There are four major takeaways from these results. First, in the oblasts of southeast Ukraine, there is general skepticism about political leaders, with only Poroshenko receiving a small positive likeability rating among Ukrainians who said they spoke both Russian and Ukrainian at home. Putin received by far the highest negative ratings of the four leaders. In the smaller Crimea survey, Putin scored strongly on the likeability scale among both Russians and Ukrainians, with Tatar respondents on balance neutral. Aksyonov scored moderately well in the likeability scale of this overwhelmingly Russophonic sample. Obama’s rating is a mirror image of Putin’s in Crimea but even in the six southeast oblasts, the overall score is negative for all groups.

Second, there are no dramatic differences between ethnic Russia and ethnic Ukrainian opinion in both survey locations. Neither group is politically homogeneous, and both feature a range of opinions. While it is noteworthy that ethnic Russians are likely to dislike Obama and like Putin more, a plurality of ethnic Russians in the sampled sites in southeast Ukraine dislike Putin more than like him. There are also minor differences between ethnic Ukrainians and Russians in Crimea but the salient divide there is between the Tatar population and the rest. Tatar respondents recorded the strong dislike of Lenin across the sites (no doubt explained by their appalling treatment by the Soviet state including their deportation from the peninsula in 1944 to Central Asia) and revealed the weakest levels of support for Putin in the Crimea context.

Third, it is superficial to claim that Ukraine is divided along ethnicity and language lines. These results reveal that this is not the case. Politicians and pundits often conflate language and nationality in Ukraine but as the preliminary results on other questions in our survey show, there is little correlation between the strength of feelings for the Ukrainian nation and language use in the southeast of the country.  Leading scholars argue correctly that Ukraine is a state of historically distinctive regions. The particular spatial context matters a great deal in the formation of political opinion. Clearly there are big differences from the Galician oblasts of the west to the Russified oblasts of the east in electoral preferences and support for joining Western institutions like NATO or the European Union. This does not mean sharply divergent levels of allegiance to Ukraine.

Finally it is worth emphasizing that the internal regional preferences do not explain the current Ukraine crisis. Ukraine is divided today by the fact that Russia has outright annexed one of its regions. The majority population there supports this move despite widespread international condemnation of the annexation as illegal and illegitimate. Beyond the separatist Donetsk and Lugansk regions that Russia supports but within so-called “Novorossiya,” Putin’s actions have alienated a clear majority of Ukrainians, whether they are ethnic Ukrainians or ethnic Russians, Ukrainophones, Russophones or the many who speak both languages.  Other responses in our survey there reveal that only 5 percent of Ukrainians have ‘bad feelings’ about Russians but the same certainly cannot be said for the Russian leader.

Gerard Toal (Gearóid Ó Tuathail) is Director of the Government & International Affairs program at Virginia Tech’s National Capital Region campus in Old Town, Alexandria, Virginia. John O’Loughlin is College Professor of Distinction and Professor of Geography at the University of Colorado, Boulder.

Edit: The figures were corrected to replace the original label “Eastern Ukraine” with “6 oblasts SE Ukraine”