The results of the latest Eurobarometer poll on the Ukrainian war can be explained on only two axis: overall satisfaction and refugees acceptance. An exercise in clickbait.

The most recent survey of the public opinion in the European Union contains a dozen questions directly pertaining to the war in Ukraine. The poll tries to gauge the satisfaction with the actions taken by the EU in support of Ukraine. The respondents are asked to indicate their level of satisfaction with such actions as “Financing the purchase and supply of military equipment to Ukraine” or “Welcoming into the EU people fleeing the war”. There are also questions regarding the citizen’s agreement with the anti-Russian stance of the EU.

Overall, the europeans seem to support the pro-Ukraine position. I wanted to see if some quick-and-dirty data analysis of the raw data can uncover some interesting trends.

After extracting the data I chose to work with the question-by-country matrix containing the proportion of positive answers (the questions are such that the positive answers are always pro-Ukrainian). A low-rank (2) approximation of the matrix shows that more than two-thirds of the variation in this matrix can be explained by only two directions. Each axis could be associated with a set of priorities or issue that determine the differences in the public attitudes towards the war.

Two main issues (axes) that explain the difference between countries. Each ‘issue’ contribute to the poll result of a question with the corresponding weight. The first, being uniformly positive for all the questions, could be termed “overall satisfaction with pro-Ukraine policies”. As much as half of the differences between the countries can be explained by a single number, quantifying the strength of pro-Ukraine (or anti-Russian) sentiment, applied uniformly to every aspect of the war. The second issue is negatively associated with “supporting the refugees”. Indeed, question 5 deals with humanitarian support, 6 with receiving refugees and 7 with direct financial support for Ukraine. Alternatively, the second issue is negatively related to the visible, immediate financial burden.

The plot above contains a projections of the answers (%) to eleven questions on the 2D surface of the two composite axes supporting most of the variation. The horizontal axis is the variation of the average of the overall satisfaction with EU policies and agreement with their importance. For me is somewhat surprising that the central-south-european countries seem to be, uniformly, the least anti-russian. I can understand the Bulgarians, due to their cultural ties, but I was hoping that the Romanians would remember the not-so-distant past! The level of support shown in Poland, baltic and scandinavian states is not really surprising. As an aside I am always surprised how different are the Danish from the Germans.

The other surprise is that the second most important source of variation (the vertical axis above) is strongly defined by the level of support for the Ukrainian refugees. OK, the Romanians can be quite nasty, but why would be the Greeks so accepting of refugees: don’t they have enough problems with people fleeing the wars in the middle-east?