In airstrike-only third-party military interventions, what are the consequences
bore on civilians? Prior studies have worked towards answering this question but
lacked a “clean case” where only two warring groups were intervened upon. The
lack of “clean cases” provides noise in making causal estimates between airstrikes
and civilian victimization. I fill this gap by exploring the Kosovo War, where
NATO intervened with airstrikes against the Yugoslav forces who were fighting the
KLA guerrillas. I link high-resolution spatiotemporal data on civilian victimization
to a novel dataset I created with all known NATO airstrikes to investigate this
relationship. I find that levels of civilian victimization increased shortly after the
introduction of strikes and following a shift in strike strategy. Yet, I also find
that while strikes at-large increased civilian victimization, in municipalities recently
struck, they decreased the levels of civilian victimization. However, strikes also
increased the expected count of civilian victimization following battles and troop
losses. Therefore, while strikes were effective in municipalities recently struck, I find
they increase the expected levels of civilian victimization through other means. This
study shows the implications of airstrike-only interventions and has implications on
future policy and military strategy.