EU member state participation in military operations: a configurational comparative analysis

Abstract

Contrary to what the EU’s image as a civilian power suggests, the EU member states have had 50,000–100,000 troops deployed outside their home countries for most of the post-Cold-War period. Although the vast majority of these troops were active in operations with a strong European presence, the member states’ patterns of military engagement differ significantly. This study provides a systematic analysis of the member states’ contributions to military operations. More specifically, it examines which (combinations of) conditions consistently led to participation in EUFOR Congo, UNIFIL II, EUFOR Chad, the 2011 military intervention in Libya and the air strikes against the Islamic State. Methodologically, it builds on most different similar outcome/most similar different outcome and qualitative comparative analysis. The results of the analysis show that four conditions account for the bulk of the member states’ patterns of military engagements: military resources, competing deployments, UN peacekeeping tradition and regional trade.

Notes

¹ The member state contributions to UNIFIL II, EUFOR Congo and EUFOR Chad have been examined in prior studies (Haesebrouck 2015; Haesebrouck and Thiem Forthcoming); while some of the member states were included in studies on North Atlantic Treaty Organization (NATO) burden sharing in Libya and democratic participation in the air strikes against IS (Haesebrouck 2016b; 2016c). However, the present study constitutes the first analysis that aims to explain the pattern of participation across the five operations.

² EUFOR Congo was deployed to support the United Nations Mission in the Democratic Republic of the Congo (MONUC) during the 2006 presidential election; EUFOR Chad was tasked with the protection of the personnel of the United Nations Mission in the Central African Republic and Chad (MINURCAT) and was intended as a bridging mission whilst the force generation took place for the military component of MINURCAT.

³ According to the IISS (2014) military balance, Estonia, Ireland, Latvia, Lithuania, Luxembourg and Slovenia do not have fighter jets.

⁴ The coding of the operation against IS reflects the situation in 2014. In November 2015, the German cabinet decided to participate in the operation after terrorist attacks in Paris and the subsequent adoption of UN Resolution 2249, which provided a legal basis for participation (Muller-Neuhof 2015).

⁵ The operationalization of some of the conditions in some operations builds on prior research (see Haesebrouck 2015, 2016b, 2016c; Haesebrouck and Thiem Forthcoming).

⁶ The results were produced with the MDSO/MSDO software (version 1.1 – spring 2015), developed by G De Meur (available at http://www.jchr.be/01/v11.htm) Replication data are available in ‘replication data MDSO MSDO.txt’ (Supplementary material).

⁷ In line with the hypotheses above, large and medium military expenditures, large and medium military personnel, geographic proximity, bilateral and regional trade, foreign fighters, alliance value, peacekeeping tradition and distant elections were linked to military participation, as was the absence of large share sustainable forces deployed, above average competing deployments, budget deficit, parliamentary veto power, minority government, constitutional restrictions and coalition government. No directional expectations were made on the conditions related to ideology, since their impact is dependent on the context of the operation.

⁸ The systematic cross-case comparison is available in ‘comparison MDSO MSDO.xlsx’.(Supplementary material)

⁹ Moreover, the analyses presented in Appendix 4 demonstrate that severe model ambiguities arise if models with seven conditions are tested, suggesting that a larger number of conditions would result in very ambiguous results.

¹⁰ The R code is provided in ‘code.R’, replication material in ‘replication data QCA.txt’ and ‘replication data QCA cases IS.txt’ (Supplementary material). The analysis of necessity, which is presented in appendix 5, shows that there are no non-trivial necessary conditions for either the outcome or the absence of the outcome.

¹¹ On top of that, as shown in the discussion of the contradictory truth table rows, the latter generally include cases that are best explained by idiosyncratic or case-specific circumstances that cannot be captured in conditions that are relevant and are generalizable to the other cases (see also ‘Discussion’ below).

¹² The solution for military participation explains the outcome in 38 of the 48 cases (coverage of 0.792), the solution for ~military participation explains the absence of the outcome in 43 of the 61 cases (coverage of 0.705). In total, the two solutions thus explain 81 of the 109 cases or 74.3 per cent.

¹³ The cross-case comparisons are available in ‘contradictory cases IS.xlsx (Supplementary material)’.

¹⁴ The conservative and intermediate solutions are presented in Appendix 2.