From reluctance to reassurance: Explaining the shift in the Germans’ NATO alliance solidarity following Russia’s invasion of Ukraine

ABSTRACT

Russia’s invasion of Ukraine in 2022 has led to a shift of public opinion in Germany on collective defense: reluctance towards the defense of NATO’s eastern flank has given way to majority support for military efforts to reassure NATO’s eastern members in the face of Russian aggression. Against the background of the war, which factors are driving this shift in peoples’ alliance solidarity? An in-depth analysis of representative population surveys from 2021 and 2022 shows that the perception of Russia as a security threat is the key factor, yet it is only part of a more complex explanation involving peoples’ strategic postures and issue-specific knowledge as well. Contrary to common belief, a free-riding mentality turns out to be irrelevant. The empirical findings shine light on Germany’s reaction to Russia’s war against Ukraine and add to our understanding of the societal foundations of alliance solidarity in Germany and other NATO countries.

KEYWORDS:

• NATO
• Germany
• Russia
• public opinion
• strategic postures
• threat perception

Acknowledgements

The authors would like to express their gratitude to the two anonymous reviewers for their extensive, yet very precise feedback, which has greatly helped to improve both the analysis and the argument.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Please note that we do not assume the explanatory factors to work as a causal chain. Instead, each explanatory factor is expected to exhibit a statistically significant association with alliance solidarity on its own. Hence, we do not test for mediation effects.

2 The 2022 dataset contains four factual knowledge questions on collective defense in the context of NATO. These questions were not asked in 2021. Since the goal of the analysis is to describe and explain changes over time the factual knowledge questions are not included as an independent variable in the main analysis. The four factual knowledge questions in 2022 read: (1) Which of the following countries is not a NATO member? (2) Which institution needs to approve the deployment of the Bundeswehr on NATO territory? (3) Germany participates currently in NATO’s Enhanced Forward Presence in Lithuania. Which branch of the Bundeswehr is in charge of that mission? (4) To which of the following countries has the Bundeswehr never been deployed? (featuring some former NATO mission countries and one incorrect response option). The number of correct answers can be summed up for every participant as an index of their factual knowledge. When the participants’ factual knowledge (instead of the subjective knowledge) is included as an independent variable in the 2022 multiple regression analysis, we find two things. First, the effect sizes of the other independent variables do not change, which testifies to the robustness of the overall model (for the regression results see Table S1 as supplementary material online). Second, the standardized effect of factual knowledge on the participants’ alliance solidarity is half the size of the participants’ subjective knowledge, meaning subjective knowledge is more strongly associated with alliance solidarity than factual knowledge. However, the amount of explained variance in the participants’ alliance solidarity does not change from the model presented in the main analysis. Furthermore, the participants’ factual knowledge correlates significantly with their subjective knowledge about the Bundeswehr’s participation in EFP Lithuania and the Baltic Air Policing (r = .31; p < .001). Based on these observations, we argue that the participants’ level of subjective knowledge about the Bundeswehr missions on NATO’s eastern flank is an adequate (yet imperfect) proxy for the participants’ factual knowledge.

3 The execution of an exploratory factor analysis (principal components analysis) results in the extraction of a single factor in both years of analysis: 2021 (Eigenvalue = 2.17; explained variance = 54 percent); 2022 (Eigenvalue = 2.54; explained variance = 63 percent). All items have loadings larger than .7. These findings support the unidimensional operationalization of the dependent variable “alliance solidarity.”

4 The execution of an exploratory factor analysis (principal components analysis) results in the extraction of a single factor in both years of analysis: 2021 (Eigenvalue = 2.62; explained variance = 65 percent); 2022 (Eigenvalue = 2.86; explained variance = 71 percent). All items have loadings larger than .7. These findings support the unidimensional operationalization of the perceived threat from Russia.

5 The execution of an exploratory factor analysis (principal components analysis) results in the extraction of a single factor in both years of analysis: 2021 (Eigenvalue = 2.04; explained variance = 68 percent); 2022 (Eigenvalue = 2.00; explained variance = 65 percent). All items have loadings larger than .7. These findings support the unidimensional operationalization of the strategic posture Militarism.

6 The execution of an exploratory factor analysis (principal components analysis) results in the extraction of a single factor in both years of analysis: 2021 (Eigenvalue = 1.79; explained variance = 60 percent); 2022 (Eigenvalue = 1.95; explained variance = 65 percent). All items have loadings larger than .7. These findings support the unidimensional operationalization of the strategic posture Atlanticism.

7 We also estimated a more complex model including socio-demographic characteristics and additional attitudinal variables. However, these variables are mostly irrelevant for explaining the participants’ alliance solidarity. Moreover, the inclusion of these control variables has yielded almost identical results for our explanatory variables of interest.

8 We also estimated all regression models with the remaining combinations of the attitude on alliance membership (member) and the willingness to honor financial obligations to NATO (pay) as possible reference groups for the free-rider dummy: member/pay; no member/no pay; no member/pay. These analyses generated results for the free-rider dummy that are very similar to the ones presented here.

9 We thank one of the anonymous reviewers for bringing up this argument in his/her review.

Additional information

Notes on contributors

Timo Graf

Timo Graf is a senior researcher in the Department of Military Sociology at the Bundeswehr Center of Military History and Social Sciences in Potsdam, Germany. He is the leading project manager of the annual population survey conducted on behalf of the German Ministry of Defense. He specializes in quantitative research on public opinion on foreign, security, and defense policy as well as on civil-military relations. Prior to his service with the armed forces, he was part of an international research project of the European Commission on Chinese public opinion on China-EU relations.

Markus Steinbrecher

Markus Steinbrecher is a senior researcher in the Department of Military Sociology at the Bundeswehr Center of Military History and Social Sciences in Potsdam, Germany. His research focuses on political behavior, public opinion, and military sociology. He has published widely on political participation, economic voting, political attitudes on foreign and security policy, and civil-military relations. Before joining the Bundeswehr, he held positions at the University of Mannheim, Northwestern University in Evanston, IL, US, and the Otto-Friedrich-University, Bamberg.

Heiko Biehl

Heiko Biehl is a senior researcher in the Department of Military Sociology at the Bundeswehr Center of Military History and Social Sciences in Potsdam, Germany. He specializes in military sociology, in public opinion, and in strategic and military culture. He has published widely in the fields of military sociology, political sociology, and security policy. Previously, he held positions at the University of Potsdam, the Bundeswehr Academy for Information and Communication in Strausberg, and the Bundeswehr Command and Staff College in Hamburg.