Table 14.

Number of new locations in district.

Panel A
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
Conditions variable:	Area of district	Total population	Population density	Nighttime lights	NDVI	Urban share	Cropland share
Crossed	−12.58	−12.77	−12.36	−12.76	−11.78	−12.50	−14.20
threshold	(12.75)	(12.89)	(12.58)	(12.29)	(7.805)	(12.13)	(12.42)
Conditions	−0.941	−0.904	−0.000363	0.0640	−0.00343	10.99	1.153
	(0.947)	(0.660)	(0.000299)	(0.0764)	(0.00195)	(12.34)	(4.567)
Crossed	1.017	1.346	0.000878	0.145	−0.0000630	14.72	4.707
threshold = 1 $\times$ Conditions	(1.032)	(1.152)	(0.000723)	(0.124)	(0.00185)	(22.06)	(2.942)
Observations	104960	104960	104960	104960	104960	104960	104960
Panel B
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Conditions variable:	Road dist. to cities	Dist. to coast	Dist. to waterway	Dist. to petroleum	Dist. to gold	Dist. to diamonds	Conflict fatalities	Princ. Comp.
Crossed	−12.96	−13.00	−13.97	−11.42	−17.40	−16.98	−12.82	−12.17
threshold	(14.67)	(16.00)	(16.38)	(12.81)	(17.59)	(17.11)	(12.67)	(12.58)
Conditions	−0.00701	−0.00461	−0.0389	0.0150	−0.00437	−0.00129	0.0849
	(0.00910)	(0.00594)	(0.0790)	(0.00938)	(0.00773)	(0.00289)	(0.316)
Crossed	0.00249	0.00236	0.0444	−0.00246	0.0156	0.00732	43.16***
threshold = 1 $\times$ Conditions	(0.00678)	(0.00985)	(0.101)	(0.00527)	(0.0156)	(0.00729)	(2.537)
PC 1								1.072
								(1.748)
PC 2								0.0557
								(1.726)
PC 3								−1.513
								(1.744)
Crossed								0.675
X PC1								(1.777)
Crossed								1.846
X PC2								(3.456)
Crossed								−0.0563
X PC3								(0.951)
Observations	104900	104960	104960	104960	104960	104960	103500	103500

Standard error clustered by country and year in parentheses. *** $p<0.01$, ** $p<0.05$, * $p<0.1$.