Voters’ turnout ratio is the ratio of valid votes on all registered voters in this referendum, and the results show this referendum voters’ turnout ratio is only 35.3%, which is especially low (see table 1). Analyzing whether rebel violence influences voters’ willingness to vote might explain the reason why this turnout ratio is so low.
The total rebel violence for both characteristics (civilian killings and kidnappings) show no significant relationship with voters’ turnout ratio (see figure 4).
Civilian killings on municipal level shows there is no relationship between civilian killings and voters’ turnout (see figure 5). However, the data shows a slightly negative
relationship between civilian killings on per capita level and voters’ turnout (see figure 6). Since civilian killings on per capita level shows an average effect among all population instead of one municipality, so per capita level should show the effect more precisely. Therefore, citizens who exposed to civilian killings before are more likely to not show up in the peace accord referendum.
Both kidnappings on municipal level and on per capita level show no significant relationship between voters’ turnout and kidnappings (see figure 7), this means citizens who exposed to kidnappings before have no impact on the voters’ turnout ratio.
In general, average rebel violence ratio does not impact voters’ turnout ratio, yet some specific characteristics can. Citizens who exposed to kidnappings before does not influence the voters’ turnout ratio, however, citizens who exposed to civilian killings before shows to turnout less during the referendum. Civilian killings could possibly depress the person’s willingness to participate activities in public, therefore people who exposed to civilian killings are tend to less likely to show up during the referendum.