Sure would be nice if social science were complete, and we could say with scientific certainty the effect of any given social activity.
Here are a couple published articles that attempt to get at that question, more research required tho.
https://journals.sagepub.com/doi/full/10.1177/07255136231154266
https://journals.sagepub.com/doi/full/10.1177/00936502211032100
Give the Kurds them both you cowards. Damn the expense!