Editing Civil war (section)

===== Effectiveness of intervention =====
The effectiveness of intervention is widely debated, in part because the data suffers from selection bias; as Fortna has argued, peacekeepers select themselves into difficult cases.<ref>{{Cite book|title=Does peacekeeping work? shaping belligerents' choices after civil war|last=Fortna|first=Virginia Page|date=2008|publisher=Princeton University Press|isbn=9780691136714|oclc=785583130}}</ref> When controlling for this effect, Forta holds that peacekeeping is resoundingly successful in shortening wars. However, other scholars disagree. Knaus and Stewart are extremely skeptical as to the effectiveness of interventions, holding that they can only work when they are performed with extreme caution and sensitivity to context, a strategy they label 'principled incrementalism'. Few interventions, for them, have demonstrated such an approach.<ref>{{Cite book|title=Can intervention work?|last=Gerald.|first=Knaus|isbn=9780393342246|oclc=916002160|year=2012|publisher=National Geographic Books }}</ref> Other scholars offer more specific criticisms; Dube and Naidu, for instance, show that US military aid, a less conventional form of intervention, seems to be siphoned off to paramilitaries thus exacerbating violence.<ref>{{Cite journal|last=Dube, Vargas|date=2015|title=Bases, Bullets and Ballots; The Effect of US-Military Aid on Political Conflict in Colombia|journal=The Journal of Politics|volume=77|issue=1 |pages=249–267|doi=10.1086/679021|citeseerx=10.1.1.622.2394|s2cid=220454361}}</ref> Weinstein holds more generally that interventions might disrupt processes of 'autonomous recovery' whereby civil war contributes to state-building.<ref>{{Cite web|url=https://www.cgdev.org/files/2731_file_WP57.pdf|title=Autonomous Recovery and International Intervention in Comparative Perspective, CGDEV Working Paper|last=Weinstein|first=Jeremy|date=April 2005|archive-url=https://web.archive.org/web/20120528160735/http://cgdev.org/files/2731_file_WP57.pdf|archive-date=2012-05-28|url-status=dead}}</ref>

On average, a civil war with interstate intervention was 300% longer than those without. When disaggregated, a civil war with intervention on only one side is 156% longer, while when intervention occurs on both sides the average civil war is longer by an additional 92%. If one of the intervening states was a superpower, a civil war is a further 72% longer; a conflict such as the [[Angolan Civil War]], in which there is two-sided foreign intervention, including by a superpower (actually, two superpowers in the case of Angola), would be 538% longer on average than a civil war without any international intervention.{{sfn|Hironaka|2005|pp=50–51}}