Crime and immigration: new evidence from England and Wales

Table 5 Robustness checks

	Annual Population Survey, 2004/5-2010/11
	Change in log crime rate		Change in crime rate,		Change in burglary rate,
	Change in log crime rate		Crime dynamics		Burglary dynamics
	First stage	Second stage	First stage	Second stage	First stage	Second stage
	(1)	(2)	(3)	(4)	(5)	(6)
A. England and Wales
Predicted change in immigrant Share	0.586 (0.162)		0.555 (0.099)		0.558 (0.097)
High historical immigrant share	-0.158 (0.045)		-0.055 (0.0145)		-0.055 (0.014)
Change in immigrant share		0.008 (0.060)		-0.118 (0.101)		0.011 (0.017)
Change in crime/burglary rate, 2002-2004				0.475 (0.124)		0.292 (0.042)
Controls	Yes	Yes	Yes	Yes	Yes	Yes
Police force area fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
F-test	14.56		18.65		18.54
Sample size	347	347	347	347	347	347
B. London
Predicted change in immigrant share	0.761 (0.460)		0.726 (0.162)		0.671 (0.136)
High historical immigrant share	-0.159 (0.050)		-0.055 (0.017)		-0.048 (0.020)
Change in immigrant share		0.019 (0.183)		-0.080 (0.056)		0.010 (0.017)
Change in crime/burglary rate, 2002-2004				0.182 (0.021)		0.266 (0.074)
Controls	Yes	Yes	Yes	Yes	Yes	Yes
Police force area fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
F-test	10.04		13.70		10.21
Sample size	33	33	33	33	33	33

Notes: Weighted by population. High Historical Immigrant Share is a dummy variable equal to one if the Immigrant Share in 1991 Census of the local authority is greater than 0.20. Controls are: population growth, the change in the unemployment rate, the change in the share of males aged 15–39 and a dummy variable for the 33 London boroughs. Robust standard errors in parentheses. In columns 3 and 5 there are two instrumented endogenous variables and the Angrist-Pischke multivariate F-Test of excluded instruments is reported. Burglary and Crime rates are burglary and crime counts divided by total population from the APS.