Table 6 Heterogeneous responses by quartile of violence
From: Violence and migration: evidence from Mexico’s drug war
Quantiles predicted homicides without highways | ||||||||
|---|---|---|---|---|---|---|---|---|
Panel A: domestic migration | Quartile = 1 | Quartile = 2 | Quartile = 3 | Quartile = 4 | ||||
Municipal cross section | First stage | Second stage | First stage | Second stage | First stage | Second stage | First stage | Second stage |
(1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | |
Federal highway kilometers 2005 | 0.11674 | 0.1578*** | 0.4082*** | 0.45109** | ||||
(0.084) | (0.061) | (0.065) | (0.192) | |||||
Homicides 2007–2010 | 0.02620 | 0.01700 | −0.00908 | 0.01125 | ||||
Per 100,000 inhabitants | (0.045) | (0.020) | (0.009) | (0.011) | ||||
Observations | 606 | 606 | 607 | 607 | 607 | 607 | 605 | 605 |
Angrist-Pischke F test | 1.915 | 1.915 | 6.765 | 6.765 | 39.29 | 39.29 | 5.518 | 5.518 |
Kleinbergen-Paap rK LM ChiSq. | 1.941 | 1.941 | 6.802 | 6.802 | 37.48 | 37.48 | 5.559 | 5.559 |
Panel B: international migration | Quartile = 1 | Quartile = 2 | Quartile = 3 | Quartile = 4 | ||||
Municipal panel | First stage | Second stage | First stage | Second stage | First stage | Second stage | First stage | Second stage |
(1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | |
Federal highway kilometers 2005* | 0.08412 | 0.07718 | 0.4501*** | 0.94848*** | ||||
Cocaine seizures in Colombia | (0.069) | (0.066) | (0.094) | (0.198) | ||||
Homicides per 100,000 inhabitants | −0.01388 | −0.01768 | −0.00347 | −0.00327 | ||||
(0.017) | (0.019) | (0.003) | (0.002) | |||||
Observations | 2351 | 2351 | 2378 | 2378 | 2383 | 2383 | 2400 | 2400 |
Angrist-Pischke F test | 1.502 | 1.502 | 1.374 | 1.374 | 23.05 | 23.05 | 22.91 | 22.91 |
Kleinbergen-Paap rK LM ChiSq. | 1.510 | 1.510 | 1.381 | 1.381 | 22.96 | 22.96 | 22.82 | 22.82 |