Working Paper: CEPR ID: DP5715
Authors: William Easterly; Roberta Gatti; Sergio Kurlat
Abstract: Using a newly assembled dataset spanning from 1820 to 1998, we study the relationship between the occurrence and magnitude of episodes of mass killing and the levels of development and democracy across countries and over time. Mass killings appear to be more likely at intermediate levels of income and less likely at very high levels of democracy. However, the estimated relationship between democracy and probability of mass killings is not linear in the full sample. In the 20th century, discrete improvements in democracy are systematically associated with episodes involving fewer victims.
Keywords: democracy; growth; mass killings
JEL Codes: N40; O10
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| mass killings (H84) | likelihood of mass killings (H84) |
| income (E25) | mass killings (H84) |
| democracy (D72) | mass killings (H84) |
| democracy (D72) | likelihood of mass killings (H84) |
| intermediate levels of income (D31) | mass killings (H84) |
| perfect democracy score (D72) | lower probability of mass killings (D74) |
| third quartile democracy (D72) | frequency of mass killings (H84) |
| 19th century wealth (N93) | mass killings (H84) |
| 20th century poverty (I32) | mass killings (H84) |