Abstract
Although a great deal of attention is paid to reproductive health during violent conflicts, the literature is sparse on the consequences of conflict for abortion and miscarriage. This research provides an analysis of a recent historical case: the 1992-1997 civil war in Tajikistan, using the female questionnaire of the 2007 Tajik Living Standards Survey to examine a subsample of 1,445 women surveyed who had reached menarche during or after the war and had been pregnant at least once by the time of the survey. The analysis leverages the uneven geographical scope of conflict events during the civil war to pinpoint women's exposure to violence, measured by the Uppsala Conflict Data Program. The results show that for women who had reached menarche during or after the civil war, exposure to conflict events increases the likelihood of ever experiencing miscarriage, but not abortion. Including a spatial lag operator reveals that there were also spillover effects for abortion, in which women who were in a broader region of uncertainty were more likely to induce an abortion. These findings highlight the role of institutional changes in affecting pregnancy loss during and after civil war.