Non-UN Peacekeeping Effectiveness: Further Analysis

ABSTRACT

This paper examines the effectiveness of non-UN-led peacekeeping operations (PKOs) from two alternative perspectives. First, the four kinds of regional and international (out of region) PKOs are investigated based on their ability to curtail one-sided violence (OSV) against civilians by host governments or rebels. That analysis is further bolstered by propensity-score matching to ameliorate potential selection bias stemming from non-UN PKOs. For the matched sample, we find that non-UN peacebuilding and peace enforcement missions limit rebel caused OSV, which is a novel result. Second, the ability of non-UN PKOs’ troops and police to end conflict or to maintain peace are ascertained based on survival analysis. Non-UN troops, but not police, curtail the transition from peace, regardless of matching. Generally, non-UN PKOs display effectiveness in limiting OSV or maintaining peace, but not in ending conflict.

KEYWORDS:

• Non-UN PKO effectiveness
• one-sided violence
• survival analysis
• propensity-score matching

 :

• JEL Codes D74
• H56

Acknowledgments

We thank two anonymous reviewers and Khusrav Gaibulloev for helpful comments on an earlier draft. Full responsibility rests with the authors.

Disclosure Statement

No potential conflict of interest was reported by the authors.

Notes

^1. On the four kinds of UN PKOs, see, e.g., Diehl (1993), Doyle and Sambanis (2000), Fortna (2004, 2008), and Kim, Sandler, and Shimizu (2020).

^2. We obtain these mandates from the missions’ websites such as those previously cited.

^3. Inclusion of these variables does not change key results and are available upon request.

^4. 1993.1 indicates the first month of 1993. Other dates are read the same way.

^5. The absence of right censoring can occur if the conflict ends in December 2016 or if the conflict ends within 24 months before December 2016. The latter applies because we keep track of post-conflict peace for 24 months, following the method of Bara and Hultman (2020).

^6. See Kim, Sandler, and Shimizu (2020) for a more detailed discussion of accounting for multiple transitions.

^7. Some regions – e.g., South Asia – are left out because of a lack of regional PKOs.

^8. These values follow from the following calculations: $0.089-1=-0.911,$ $0.029-1=-0.971,$and $0.118-1=-0.882$.

^9. This follows from $exp\left[ln\left(1.209\right)\times ln\left(1.01\right)\right]-1=0.00189$. Other calculations involving the logs of independent variables are computed in a similar fashion, but not displayed.