Undocumented	immigrants,	U.S.	Citizens,	and	Convicted	Criminals	in	
Arizona*	
	
	
	

	
John	R.	Lott,	Jr.	
President	
Crime	Prevention	Research	Center	
johnrlott@crimeresearch.org	
(484)	802-5373	
	
	
	
	
December	1,	2017	
Revised	January	17,	2018	
Revised	February	3,	2018	
Revised	February	4,	2018	
Revised	February	10,	2018	

	
	
	
	
	
	
	
	
	
	
	
	
*	Michael	Block,	Edgar	Browning,	Tim	Groseclose,	Maxim	Lott,	Jose	Marichal,	Alex	Nowrasteh,	Paul	Rubin,	
Robert	VerBruggen,	and	Mike	Weisser	provided	helpful	comments.		In	particular,	Bill	Montgomery,	who	
represented	the	Arizona	Prosecuting	Attorneys’	Advisory	Council	and	is	the	DA	for	Maricopa	County,	was	
very	helpful.		Rujun	Wang	and	Roger	Lott	provided	helpful	research	assistance.		But	none	of	these	people	
bear	responsibility	for	the	final	product.

Electronic copy available at: https://ssrn.com/abstract=3099992

Undocumented	immigrants,	U.S.	Citizens,	and	Convicted	Criminals	in	
Arizona	
	
	
Summary	
	
■ Based	on	data	from	1985	to	2017,	undocumented	immigrants	are	at	least	146%	
more	likely	to	be	convicted	of	crime	than	other	Arizonans.			
■ Undocumented	immigrants	tend	to	commit	more	serious	crimes	and	serve	
10.5%	longer	sentences	than	do	U.S.	citizens.	
■ Undocumented	immigrants	are	163%	more	likely	to	be	convicted	of	1st	degree	
murder	than	are	U.S.	citizens,	168%	more	likely	to	be	convicted	of	2nd	degree	
murder,	and	189.6%	more	likely	to	be	convicted	of	manslaughter.		Those	three	
categories	and	negligent	homicide	added	up	to	987	incarcerations.		
Undocumented	immigrants	are	also	much	more	likely	to	commit	sexual	offenses	
against	minors,	sexual	assault,	DUI,	and	armed	robbery.		
■ Young	convicts	are	especially	likely	to	be	undocumented	immigrants.		
Undocumented	immigrants	born	after	June	15,	1981	are	eligible	for	Deferred	
Action	for	Childhood	Arrivals	(DACA).		While	undocumented	immigrants	from	15	
to	35	years	of	age	make	up	a	little	over	two	percent	of	the	Arizona	population,	
they	make	up	almost	8%	of	the	prison	population.		These	immigrants	also	tend	
to	commit	more	serious	crimes.	
■ Undocumented	immigrant	criminals	are	45.4%	more	likely	than	other	criminals	
to	have	been	gang	members,	and	133%	more	likely	to	receive	sentencing	
enhancements	for	being	classified	as	dangerous.	
■ These	numbers	may	give	an	artificially	low	estimate	of	the	share	of	crime	
committed	by	undocumented	immigrants.	Undocumented	immigrants	may	
commit	many	of	their	crimes	against	each	other,	and	their	victims	may	be	afraid	
of	calling	the	police	or	testifying	at	trial	because	of	their	undocumented	status.	
■ While	undocumented	immigrants	are	more	likely	to	be	convicted	of	crimes,	they	
also	exhibit	extremely	low	recidivism	and	criminal	history	rates.		Among	
criminals	who	are	U.S.	citizens,	a	small	subset	keeps	going	in	and	out	of	prison.		
Among	undocumented	immigrants,	a	much	larger	proportion	go	to	prison	once	
or	twice	and	then	never	return	to	prison.		24.8%	of	U.S.	citizens	were	admitted	
five	or	more	times	in	the	Arizona	Department	of	Corrections,	but	that	same	
number	is	only	2.95%	for	undocumented	immigrants.		The	evidence	suggests	
that	these	individuals	leave	Arizona	after	being	incarcerated.	
■ Evidence	is	provided	for	whether	changes	in	border	enforcement	can	explain	
changes	in	undocumented	immigrants	share	of	newly	incarcerated	inmates.	
■ If	undocumented	immigrants	committed	crime	nationally	as	they	do	in	Arizona,	
in	2016	they	would	have	been	responsible	for	over	1,000	more	murders,	5,200	
rapes,	8,900	robberies,	25,300	aggravated	assaults,	and	26,900	burglaries.	

Electronic copy available at: https://ssrn.com/abstract=3099992

3
Introduction	
	
Arizona’s	prison	population	data	allow	us	to	compare	undocumented	immigrants’	share	
of	the	prison	population	with	their	estimated	share	of	the	state	population.		We	have	
data	from	the	beginning	of	1985	to	June	2017.		For	the	first	time,	we	break	down	the	
data	to	examine	differences	between	US	citizens,	undocumented	immigrants,	and	legal	
permanent	residents.			One	advantage	of	using	convictions	rather	than	just	reported	
crimes	is	that	convictions	depend	on	a	“beyond	a	reasonable	doubt”	standard	of	
evidence	and	thus	are	much	less	likely	to	count	innocent	people.	
	
Previous	research	does	not	directly	link	undocumented	immigrants	to	specific	crimes.		
After	interviewing	a	number	of	academics	who	have	done	research	in	the	area,	Politifact	
noted	in	November	2016:	"The	challenge	in	finding	concrete	numbers	is	due	to	a	
shortfall	of	data.	There	is	no	national	database	or	study	tracking	how	many	people	have	
been	killed	by	undocumented	immigrants	or	the	nationality	of	the	victims.”	Most	of	the	
literature	examines	all	immigrants,	not	just	undocumented	immigrants.		Some	studies	
on	undocumented	immigrants	depend	on	individuals	to	self-report	their	criminal	
histories	and	even	whether	they	were	born	in	the	United	States.		Others	use	proxies	for	
undocumented	immigrants	and	see	how	they	are	correlated	with	changes	in	crime	
rates.	
The	huge	advantage	of	using	the	data	that	will	be	presented	here	from	the	Arizona	
Department	of	Corrections	is	that	over	our	32.5-year	period	we	know	each	prisoner	who	
entered	the	prison	system,	their	criminal	convictions	history,	and	whether	he	is	a	
documented	or	undocumented	immigrant.		The	only	mystery	is	why	this	type	of	data	
has	not	been	utilized	until	now.	
Past	research	often	examines	rough	correlations	between	immigration	and	various	
types	of	crime	rates,	with	the	literature	divided	between	those	who	claim	immigrants	
are	more	law-abiding	and	those	who	find	no	difference	(e.g.,	Stowell	et	al,	2009,	p.	895	
for	a	survey).		Others	emphasize	more	recent	studies	that	only	find	a	benefit	in	terms	of	
lower	crime	(Landgrave	and	Nowrasteh,	2017	and	Waters	and	Pineau,	2017,	p.	326330).1		No	previous	research	over	at	least	the	last	two	decades	has	found	higher	crime	
rates	for	undocumented	immigrants	(Hagan	and	Palloni	(1998)	using	survey	data	for	
prisoners	in	El	Paso	and	San	Diego).	
Many	use	simple,	cross-sectional	analysis	to	see	whether	areas	with	higher	immigrant	
populations	have	higher	crime	rates.		Others	use	a	purely	time	series	approach.		
Rumbaut	and	Ewing	(2007)	and	Ewing	et	al.	(2015)	look	at	the	United	States	as	a	whole	
1

		Landgrave	and	Nowrasteh	(2017)	write	that	“a	vast	body	of	empirical	literature	showing	that	legal	and	
illegal	immigrants	do	not	increase	local	crime	rates,	are	less	likely	to	commit	crimes	than	their	native-born	
peers,	and	are	less	likely	to	be	incarcerated	than	are	native-born	Americans,”	but	the	literature	that	they	
cite	lumps	together	all	immigrants.	

4
and	note	that	crime	has	decreased	since	1990	as	immigration	has	increased.		They	also	
look	at	incarceration	rates	by	national	origin	and	nativity.		Stowell	et	al	(2009)	look	at	
how	crime	rates	change	in	those	metropolitan	areas	with	the	fastest	growth	in	
immigrants.			
There	are	many	different	statistical	problems	with	these	primitive	studies.		But	there	are	
also	a	number	of	data	issues	that	make	them	unable	to	infer	anything	about	the	
behavior	of	undocumented	immigrants.		Lumping	together	documented	and	
undocumented	immigrants	(and	often	naturalized	citizens)	may	mean	combining	very	
different	groups	of	people.		As	we	will	see,	documented	and	undocumented	immigrants	
have	vastly	different	incarceration	rates	in	Arizona.		Undocumented	immigrants	have	
the	highest	rates,	whereas	documented	immigrants	actually	have	lower	rates	than	do	
U.S.	citizens.		Putting	all	of	these	different	types	of	people	together,	it	is	impossible	to	
infer	anything	about	how	law-abiding	undocumented	immigrants	are.	
Other	studies	depend	heavily	on	self-reported	information,	asking	individuals	about	
their	criminal	histories	and	even	whether	they	were	born	in	the	United	States	(e.g.,	
Butcher	and	Piehl,	2007;	Ewing	et	al.,	2015;	Hickman	and	Suttorp,	2008;	Salas-Wright	et	
al.	2017).2		Undocumented	immigrants	may	not	want	to	admit	that	they	have	been	in	
prison,	fearing	that	their	criminal	record	and	illegal	status	will	make	them	prime	
candidates	for	deportation.		They	may	also	lie	about	whether	they	were	born	in	the	
United	States.	There	is	no	real	benefit	to	undocumented	immigrants	responding	
truthfully	to	the	government	or	private	surveyors.		
	
The	data	here	were	collected	for	a	report	put	together	for	the	Arizona	Prosecuting	
Attorneys’	Advisory	Council	(APAAC)	(Lott	and	Wang,	2017).		Beyond	what	criminals	are	
currently	incarcerated	for,	the	data	have	remarkable	information	on	criminal	history,	
gang	membership,	whether	they	are	identified	as	particularly	dangerous,	and	citizenship	
status.		Citizenship	status	was	determined	by	what	was	listed	in	the	pre-sentencing	
report,	and	prosecutors	and	others	knew	it	even	much	earlier	in	the	case	than	that.		This	
is	key	because	documented	immigrants	aren’t	labeled	as	“non-U.S.	citizen,	deportable”	
until	after	they	have	been	sentenced.		In	contrast,	illegal	aliens	are	labeled	that	way	
prior	to	sentencing.	
	
The	following	sections	will	first	compare	prisoners	and	the	general	Arizona	population	
by	citizenship	status,	then	by	incarceration	rates	for	younger	undocumented	immigrants	
who	are	eligible	for	Deferred	Action	for	Childhood	Arrivals	(DACA),	and	finally	by	the	
criminal	histories	of	those	in	prison.		We	then	compare	the	Arizona	prison	system	to	
what	is	available	from	the	federal	system.		Finally,	we	estimate	the	national	implications	
2

		Texas	is	one	other	state	that	provides	some	information	on	'criminal	aliens'	committing	crime,	but	that	
lumps	together	noncitizens	who	are	documented	and	undocumented	immigration	status.	See	“Texas	
Criminal	Alien	Arrest	Data,”	Texas	Department	of	Public	Safety,	checked	January	3,	2018	
(https://www.dps.texas.gov/administration/crime_records/pages/txCriminalAlienStatistics.htm).	

5
for	this	data	and	provide	information	on	how	the	results	are	affected	by	changes	in	
border	enforcement.	
	
Citizenship,	Crime,	and	Undocumented	Immigrants’	relative	shares	of	prisoners	and	
the	overall	population	
	
Undocumented	immigrants’	share	of	the	Arizona	population	appears	to	have	varied	
considerably	over	time.		Using	the	U.S.	Census,	the	Department	of	Homeland	Security	
(DHS)	estimated	that	undocumented	immigrants	made	up	2.4%,	6.39%,	and	5.48%	of	
the	state’s	population	in	1990,	2000,	and	2010,	respectively	(Figure	1).3		A	Pew	Research	
Center	analysis	of	Census	Bureau	data	estimated	a	population	share	of	4.8%	in	2014,	
and	thus	a	25-year	average	of	5.25%	from	1990	to	2014.		If	we	use	the	1990	estimate	for	
1985-1989,	and	the	2014	estimate	for	2015-2017,	then	Arizona’s	33-year	average	from	
1985-2017	would	be	about	4.8%	(though	the	trends	suggest	that	this	is	an	
overestimate).4	Using	the	PEW’s	estimates	over	the	entire	period	from	1990	to	2014	
shows	a	similar	pattern	over	time,	and	just	a	slightly	higher	average	rate	of	4.9%	over	
the	33-year	period.			

3

	DHS	put	the	number	as	high	as	8.9	in	2008,	but	the	gap	between	their	pre	and	post	census	estimates	for	
2010	was	very	large:	1.9	percentage	points.		This	suggests	that	they	overestimated	the	rate	in	2008.	
This	was	calculated	using	the	federal	government’s	estimates	for	1990,	1996,	2000,	and	2005	to	2014.		
The	values	for	the	other	years	were	filled	in	using	interpolation.	
Pew	Research	Center’s	2014	estimate	for	Arizona	available	here	(Jeffrey	S.	Passel	and	D’Vera	Cohn,	
“Overall	Number	of	U.S.	Unauthorized	Immigrants	Holds	Steady	Since	2009,”	Pew	Research	Center,	
September	20,	2016	(http://assets.pewresearch.org/wpcontent/uploads/sites/7/2016/09/31170303/PH_2016.09.20_Unauthorized_FINAL.pdf).		See	also	here	
http://www.pewhispanic.org/interactives/unauthorized-immigrants/.	
Pew	Research	Center’s	national	estimates	from	1990	to	2015	are	available	here	(Jens	Manuel,	Jeffrey	
Passel,	and	D’Vera	Cohn,	“5	facts	about	illegal	immigration	in	the	U.S.,”	Pew	Research	Center,	April	27,	
2017	(http://www.pewresearch.org/fact-tank/2017/04/27/5-facts-about-illegal-immigration-in-the-u-s/)).	
4
	This	assumes	that	the	undocumented	immigrants’	share	of	the	population	didn’t	keep	declining	as	one	
goes	further	back	in	time.	Similarly,	this	assumes	that	undocumented	immigrants’	share	didn’t	keep	
declining	after	2014.		To	the	extent	that	these	trends	had	continued	the	true	rate	would	be	even	lower	
than	the	4.8	estimate.			

6

	
	
Arizona’s	percentage	of	undocumented	immigrants	was	about	82%	above	the	national	
average	during	those	years	and	ranked	5th	in	terms	of	states	in	2014.5		The	95%	
confidence	interval	associated	with	these	estimates	is	pretty	tight,	in	2014	with	the	PEW	
ranging	from	4.54%	to	5.13%.6		
	
The	Arizona	data	show	that	undocumented	immigrants	account	for	11.8%	of	convictions	
for	first	and	second	most	serious	offenses	(11.2%	of	the	most	serious	offenses	and	
13.8%	of	the	second	most	serious	offenses).		In	2014,	the	last	year	that	we	have	data	
for,	undocumented	immigrants	accounted	for	12.6%	of	incarcerations	that	year	(11.09%	
of	the	most	serious	offenses	and	16.4%	of	the	second	most	serious	offenses).	
	
5

	Arizona’s	rank	among	states:	“Estimated	unauthorized	immigrant	population,	by	state,	2014,”	Pew	
Research	Center,	November	3,	2016	(http://www.pewhispanic.org/interactives/unauthorizedimmigrants/).	
6
	See	page	18	here	Jeffrey	S.	Passel	and	D’Vera	Cohn,	“Overall	Number	of	U.S.	Unauthorized	Immigrants	
Holds	Steady	Since	2009,”	Pew	Research	Center,	September	20,	2016	(http://assets.pewresearch.org/wpcontent/uploads/sites/7/2016/09/31170303/PH_2016.09.20_Unauthorized_FINAL.pdf.			

7
The	12.6%	share	of	2014	incarcerations	implies	that	undocumented	immigrants	were	
convicted	at	least	163%	more	often	than	Arizonans	in	general.		The	tight	confidence	
interval	associated	with	the	estimated	share	of	undocumented	immigrants	in	the	
population	would	have	to	be	over	52	standard	deviations	higher	than	it	is	for	
undocumented	immigrants	to	be	incarcerated	at	the	same	rate	as	the	average	
Arizonan.7	
	
For	the	entire	1985	to	2017	period,	undocumented	immigrants	were	146%	more	likely	
to	be	convicted.		A	useful	comparison	can	be	made	to	Hispanics.		Table	1	shows	that	in	
2014,	Hispanics	accounted	for	about	37.6%	of	entering	prison	and	about	28.8%	of	the	
Arizona	population.	That	means	their	incarceration	rate	exceeded	the	overall	state	
imprisonment	rate	by	about	31%.		(Appendix	1	breaks	down	the	data	per	person	rather	
than	per	incarceration,	though	those	results	show	a	43%	higher	share	of	undocumented	
immigrants.)	
	
In	comparison,	documented	immigrants	(Legal	Permanent	Residents)	were	extremely	
law-abiding.		They	made	up	only	1.5%	of	the	prison	population	in	2014,	while	a	rough	
estimate	indicates	that	their	share	of	the	state’s	population	is	about	3.9%.8		This	
7

One	person	suggests	that	the	Department	of	Corrections	data	mean	something	different	that	what	I	was	
told	by	the	APAAC.		“Lott	erroneously	assumed	that	the	third	category,	called	‘non-US	citizen	and	
deportable,’	only	counted	illegal	immigrants,”	claims	Alex	Nowrasteh.		He	asserts	it	also	includes	legal	
immigrants,	though	this	ignores	the	importance	of	the	pre-sentencing	report	in	collecting	this	
information.		Yet,	even	if	that	were	correct,	it	doesn’t	greatly	affect	our	results.	Only	about	10%	of	those	
deported	are	lawful	permanent	residents.		Lott	shows	that	after	also	accounting	for	temporary	foreign	
workers	this	claim	only	reduces	undocumented	immigrants’	share	of	convictions	leading	to	incarceration	
from	11.8%	to	10.6%.		Alex	Nowrasteh,	“The	Fatal	Flaw	in	John	R.	Lott,	Jr.’s	study	on	illegal	immigrant	
crime	in	Arizona,”	Cato	Institute,	February	5,	2018.		John	Lott,	“Responding	to	Cato's	and	others'	attacks	
on	our	research	regarding	crime	by	illegal	immigrants,”	Crime	Prevention	Research	Center,	February	6,	
2018	(https://crimeresearch.org/2018/02/responding-catos-attacks-research-regarding-crime-illegalimmigrants/).		
8
	The	Legal	Permanent	Resident	population	estimate	for	2014	in	Arizona	is	260,000	(James	Lee	and	Bryan	
Baker,	“Estimates	of	the	Lawful	Permanent	Resident	Population	in	the	United	States:	January	2014,”	
Department	of	Homeland	Security’s	Office	of	Immigration	Statistics	
(https://www.dhs.gov/sites/default/files/publications/LPR%20Population%20Estimates%20January%2020
14.pdf).		Refugees	and	Asylees	can	apply	to	be	legal	permanent	residents	after	one	year	of	continuous	
presence	in	the	U.S.	(Immigration	and	Refugees,	Common	Terms,	Penn	State	University	Libraries,	
http://guides.libraries.psu.edu/c.php?g=623034&p=4339995).		Add	in	refugees	and	asylees	by	assuming	
that	they	are	equally	spread	out	in	congressional	districts	across	the	country	(Refugees	&	Asylees	2015	
Data	Tables,	Homeland	Security,	https://www.dhs.gov/immigration-statistics/refugees-asylees).	
There	is	a	second	way	of	getting	at	this	number.		This	is	obtained	by	taking	the	share	of	Arizona	
population	that	are	legal	permanent	and	temporary	visas	holders	and	multiplying	it	by	the	share	of	US	
visas	holders	who	have	permanent	visas.		The	Arizona	Department	of	Corrections	numbers	are	for	
documented	immigrants	(Legal	Permanent	Residents).		The	Migration	Policy	Institute	reports	a	broader	
number	that	7.8%	(534,213)	of	Arizona’s	population	in	2015	was	“The	foreign-born	population	includes	
naturalized	U.S.	citizens,	lawful	permanent	immigrants	(or	green-card	holders),	refugees	and	asylees,	
certain	legal	nonimmigrants	(including	those	on	student,	work,	or	some	other	temporary	visas),	and	
persons	residing	in	the	country	without	authorization.”		The	Migration	Policy	Institute	also	indicates	that	

8
suggests	that	lumping	together	documented	and	undocumented	immigrants	provides	a	
very	misleading	image	of	both	groups.		However,	immigrants	as	a	group	(legal	and	illegal	
immigrants)	still	are	convicted	at	relatively	high	rates,	making	up	14.1%	of	
incarcerations	and	8.7%	of	the	population	–	an	incarceration	rate	62%	higher	than	their	
share	of	the	general	population.9		
	
There	are	some	obvious	differences	between	incarcerated	documented	and	
undocumented	immigrants	(Table	2).		Documented	immigrants	are	5.6	years	older.		Just	
34.7%	of	documented	immigrants	are	30	years	of	age	and	younger,	while	for	
undocumented	immigrants	it	is	52.4%.			In	this	dimension,	undocumented	immigrants	
are	closer	to	US	citizens	who	have	44.6%	are	in	the	same	age.		Documented	immigrants	
are	also	18.3	percentage	points	less	likely	to	be	Hispanics.		
	
If	documented	immigrants	are	so	law-abiding	and	they	continue	to	be	that	way	once	
they	become	U.S.	citizens,	including	naturalized	citizens	with	native-born	citizens	will	
make	native-born	ones	look	more	law-abiding	than	they	are.		But	the	effect	is	very	
small.		For	all	U.S.	citizens	in	2016,	they	make	up	86.9%	of	those	incarcerated	that	year	
and	92.27%	of	the	population.		Assuming	that	naturalized	citizens	are	incarcerated	at	
the	same	rate	as	documented	immigrants,	the	incarceration	rate	for	native-born	U.S.	
citizens	would	be	84.7%	and	their	share	of	the	population	86.51%.		These	numbers	
imply	only	a	3.9%	change	in	the	per	capita	incarceration	rates	for	U.S.	citizens	when	
naturalized	citizens	are	removed.	
	
Hispanics	who	are	legally	in	the	U.S.	(U.S.	citizens	and	documented	immigrants)	make	up	
26%	of	Arizona’s	population	and	29.8%	of	the	prison	population.10		But	given	that	these	
legal	Hispanics	are	very	young,	with	a	media	age	of	only	20	and	that	young	people	
generally	commit	most	crime,	adjusting	for	their	age	implies	that	they	are	law-abiding	
for	the	U.S.	as	a	whole	that	number	is	22,593,269,	and	the	U.S.	State	Department	indicates	that	for	that	
year	10,891,745	were	on	temporary	visas.		Thus,	51.8%	of	Migration	Policy	Institute	number	represents	
people	who	were	temporarily	in	the	U.S..		Assuming	that	is	the	same	rate	in	Arizona,	51.8%	of	7.8%	is	
4.04%.		The	U.S.	State	Department	annual	report	for	2016	Table	18	is	the	source	of	this	data	
(https://travel.state.gov/content/dam/visas/Statistics/AnnualReports/FY2016AnnualReport/FY16AnnualR
eport-TableXVIII.pdf).	
9
	Documented	immigrants	were	convicted	and	served	prison	time	for	a	number	of	the	most	serious	
crimes.	Twenty-five	were	convicted	of	manslaughter,	with	those	who	have	been	released	serving	an	
average	sentence	of	5.8	years.		Seventeen	of	those	went	to	prison	after	2008,	and	those	convicts	who	
have	been	released	served	an	average	of	about	2.8	years	in	prison.		The	last	two	such	people	entered	
nd
prison	entered	prison	in	2017.		Twelve	were	convicted	of	2 	degree	murder,	and	those	who	have	been	
released	served	an	average	of	13.9	years	in	prison.		Nine	of	those	entered	prison	after	2008.			Fifteen	
st
went	to	prison	for	1 	degree	murder,	and	all	of	them	had	serious	secondary	offenses	for	which	they	were	
also	convicted.		Thirteen	of	those	fifteen	went	to	prison	after	2008.		Eighteen	documented	immigrants	
were	convicted	of	sexual	assault.		
10
	Pew	Research	Center,	“Demographic	profile	of	Hispanics	in,	Arizona	2014,”	
(http://www.pewhispanic.org/states/state/az/)	and	Pew	Research	Center,	“Overall	Number	of	U.S.	
Unauthorized	Immigrants”	(https://tinyurl.com/yaaurm9r).	

9
compared	to	the	rest	of	the	legal	population.11		The	perception	that	Hispanics	are	
relatively	more	likely	to	be	criminals	is	misplaced.		It	results	from	combining	legal	and	
illegal	Hispanics.12	
	

11

	Pew	Research	Center	estimates	that	in	2014	the	median	age	for	non-Hispanic	whites	and	blacks	are	46	
and	32	respectively.		Pew	Research	Center,	“Demographic	profile	of	Hispanics	in,	Arizona	2014,”	
(http://www.pewhispanic.org/states/state/az/).	
12
	Given	that	legal	Hispanics	have	such	a	low	rate	of	convictions,	it	is	hard	to	blame	these	results	on	racism	
against	Hispanics.	

10
	
	
Table	1:	Race	and	Citizenship	of	those	incarcerated	in	Arizona	per	incarceration
2014

1985	to	June	2017
Undocumented	
immigrants:	
Non-US	citizen,	
Not	Legal	
US	
Permanent	
Citizens Resident

Documented	
immigrants:	
Non-US	
citizen,	Legal	
Permanent	
Resident

	
	
	
	
	
Total*

Undocumented	
immigrants:	
Non-US	citizen,	
Not	Legal	
US	
Permanent	
Citizens Resident

Documented	
immigrants:	
Non-US	
citizen,	Legal	
Permanent	
Resident

Caucasian 197,321

42.0%

0.6%

0.1%

8,340

39.2%

1.1%

0.1%

AfricanAmerican

61,315

12.9%

0.3%

0.1%

2,529

11.6%

0.5%

0.2%

Native	
American

29,968

6.4%

0.1%

0.0%

1,434

6.8%

0.1%

0.0%

Hispanic,	
US	Born† 120,829

25.5%

0.6%

0.0%

5,993

27.8%

1.1%

0.1%

Hispanic,	
not	US	
born
47,259

0.6%

9.2%

0.4%

1,980

1.0%

7.7%

0.9%

Asian

1,194

0.2%

0.1%

0.0%

67

0.2%

0.0%

0.1%

Other

4,950

0.5%

0.5%

0.1%

299

0.7%

0.5%

0.2%

Unknown	
race
142

0.0%

0.0%

0.0%

3

0.0%

0.0%

0.0%

Total

88.1%

11.2%

0.7%

20,645

87.4%

11.1%

1.5%

RACE

Total*

462,978

*	Totals	are	for	those	who	can	be	classified	as	U.S.	citizen,	undocumented	immigrants,	or	documented	immigrants.	0.29%	of	the	
sample	for	the	1985	to	2017	period	could	not	be	classified	in	one	of	these	three	categories,	and	0.12%	for	2014	could	not	be	
classified.	
†	One	concern	with	the	table	is	that	there	are	US	born	individuals	who	are	listed	as	not	being	US	citizens	or	who	are	attempting	to	
gain	citizenship	(though	this	last	group	is	extremely	small).		We	asked	Bill	Montgomery,	the	County	prosecutor	for	Maricopa	county,	
and	the	Arizona	Department	of	Corrections	about	these	cases,	and	we	were	told	that	they	involved	people	renouncing	their	US	
citizenship	and	then	returning	to	the	US	or	whose	US	citizenship	was	never	claimed.		Given	the	small	number	of	people	who	have	
renounced	their	citizenship,	this	assumption	seems	questionable.	(https://www.federalregister.gov/documents/2017/11/02/201723885/quarterly-publication-of-individuals-who-have-chosen-to-expatriate-as-required-by-section-6039g).		Possibly	they	are	
recording	errors,	where	non-citizens	are	listed	as	born	in	the	US	or	people	born	in	the	US	are	listed	as	non-citizens.		In	any	case,	the	
number	of	these	cases	is	very	small	and	does	not	appreciably	alter	the	results	presented	here	(See	Appendix	A3).

	

11
	
Table	2:	Demographics	of	those	incarcerated	by	Citizenship	Status	

By	Demographic	
Characteristic	
Male	
White	
Black	
NA	
Hispanic	
Asian	
Race	Other	
Race	Unknown	
Age	at	Admit	
	
Percentages	by	age	
0-20	
21-25	
26-30	
31-35	
36-40	
41-50	
51-60	
61-70	
>70	
	

US	Citizens	
86.80%	
47.66%	
14.65%	
7.26%	
29.59%	
0.21%	
0.60%	
0.03%	
33.52	
	
	
6.2%	
19.1%	
19.3%	
17.3%	
14.1%	
17.5%	
5.4%	
0.9%	
0.1%	
100.0%	

Undocumented	immigrants:	
Non-US	citizen,	Not	Legal	
Permanent	Resident	
97.30%	
4.99%	
2.42%	
0.60%	
87.15%	
0.52%	
4.29%	
0.03%	
31.43	
	
	
7.7%	
21.9%	
22.9%	
18.8%	
13.2%	
12.0%	
3.0%	
0.5%	
0.1%	
100.0%	

Documented	
immigrants:	
Non-US	citizen,	
Legal	Permanent	
Resident	
91.92%	
8.52%	
9.66%	
0.82%	
68.82%	
2.66%	
9.47%	
0.03%	
37.07	
	
	
3.4%	
14.2%	
17.1%	
14.5%	
13.5%	
23.6%	
10.8%	
2.6%	
0.3%	
100.0%	

	
The	average	prison	terms	suggest	that	undocumented	immigrants	have	committed	the	
most	serious	crimes	and	documented	immigrants	the	least	serious.		During	the	entire	
1985	to	2017	period,	the	average	prison	stay	was	660.6	days	for	undocumented	
immigrants,	598	days	for	US	citizens,	and	473.6	days	for	documented	immigrants.		
	
Possibly	undocumented	and	documented	immigrants	serve	longer	prison	sentences	
simply	because	they	are	less	likely	to	be	granted	parole	than	U.S.	citizens.		Still	that	
doesn’t	explain	the	results.		Arizona	“abolished	parole	for	offenses	committed	after	
January	1,	1994,”13	and	the	time	served	in	prison	by	undocumented	immigrants	
admitted	after	that	date	is	still	greater	than	the	time	served	U.S.	citizens:	632	days	for	
undocumented	immigrants,	553.7	days	for	US	citizens,	and	468.3	days	for	documented	
immigrants.		In	fact,	the	number	of	days	in	prison	for	undocumented	immigrants	
increases	from	being	10.5%	greater	than	U.S.	citizens	to	14.1%	greater.	
	

13

	“Truth	in	Sentencing,”	Arizona	State	Senate	Issue	Brief,	August	18,	2010	
(https://www.azleg.gov/briefs/Senate/TRUTH%20IN%20SENTENCING.pdf).	

12
The	data	here	represent	a	unique	look	at	all	of	the	prisoners	who	entered	the	Arizona	
Corrections	Department	from	January	1985	through	June	2017.		During	that	period,	
there	were	464,641	prisoners	who	entered	the	system,	and	462,978	for	whom	we	have	
information	on	their	citizenship	status.		Between	the	464,447	most	serious	offenses	and	
the	151,108	second-most	serious	offenses,	we	have	information	on	615,555	crimes.		It	is	
the	entire	universe	of	cases,	not	a	sample,	and	thus	there	are	no	issues	of	statistical	
significance.		98%	were	incarcerated	for	more	than	2	weeks.	To	the	extent	that	
differences	exist,	that	is	simply	what	the	differences	are.	
	
Tables	3,	4,	and	5	show	the	types	of	crimes	that	undocumented	immigrants	commit	at	
disproportionately	high	rates.		Table	3	provides	incarceration	rates	for	most	serious	
offenses	over	the	entire	time	for	which	we	have	data	during	the	1985	to	2017	period.		
Table	4	shows	the	same	information	for	the	second-most	serious	offense.		While	the	
results	are	generally	similar,	Table	5	matches	the	incarceration	data	for	most	serious	
offenses	in	2014	with	the	population	data	from	that	same	year.		There	were	few	
incarcerations	that	year	for	certain	types	of	crime,	so	even	a	few	cases	can	often	make	a	
big	difference	in	the	measured	incarceration	rates.		We	partially	address	this	by	limiting	
all	of	the	tables	to	showing	the	most	serious	current	offenses	for	which	at	least	20	
people	were	convicted	in	Arizona,	but	because	of	the	small	samples	the	rates	can	vary	
by	a	large	amount	from	one	year	to	another.		
	
Both	tables	show	the	high	rates	at	which	undocumented	immigrants	are	convicted	of	
serious	crime.		For	1st	Degree	Murder,	undocumented	immigrants	were	either	163%	
(Table	3)	or	232%	(Table	5)	more	likely	to	be	convicted	than	Arizonans	in	general.		For	
2nd	Degree	Murder,	they	were	either	168%	(Table	3)	or	77%	(Table	5)	more	likely,	and	
they	also	had	a	161%	higher	rate	(Table	4)	of	being	convicted	of	it	as	a	second	most	
serious	offense.14	Undocumented	immigrants	were	also	consistently	more	likely	to	be	
convicted	of	manslaughter,	armed	robbery,	sexual	assault	of	a	minor,	sexual	assault,	DUI	
or	DWI,	and	kidnapping.			
	
Given	undocumented	immigrant’s	share	of	convicted	criminals	and	the	total	crimes	that	
have	been	committed,	undocumented	immigrants	increase	total	murders	in	Arizona	
over	how	many	would	have	occurred	if	they	were	like	everyone	else	by	8.3%,	rapes	
7.97%,	robbery	3.7%,	aggravated	assault	4.3%,	burglaries	2.44%,	larceny	2.2%,	and	
vehicle	theft	3.7%.	
	
It	is	important	to	put	these	numbers	in	perspective:	relatively	few	crimes	end	up	being	
reported	or	only	a	fraction	of	those	are	solved	and	this	may	different	between	U.S.	
citizens	and	undocumented	immigrants.		First,	take	the	overall	numbers	for	robbery.		
Only	about	49.1%	were	reported	to	police	in	2014,	and	only	29.6%	of	those	that	were	
14

nd

	Combining	first	and	second	most	serious	offenses	for	2 	Degree	Murder	resulted	in	undocumented	
immigrants	facing	a	166%	higher	rate	than	their	share	of	the	population	in	committing	this	crime.	

13
reported	resulted	in	an	arrest,	implying	that	14.5%	of	all	robberies	resulted	in	arrest.15		
There	is	some	data	from	Texas	on	the	rate	that	arrest	of	undocumented	and	
documented	immigrants	result	in	conviction.		Arrests	for	robbery	over	the	period	
between	June	1,	2011	and	December	31,	2017	showed	that	51.5%	of	those	who	were	
arrested	were	convicted.16		Thus	less	than	1/7th	of	robberies	result	in	conviction,	so	the	
number	of	robbery	would	have	to	be	over	7	times	what	is	reported	here.		The	1,464	
robbery	convictions	in	our	data	from	1985	to	2017	would	imply	the	total	number	of	
robberies	was	actually	over	10,200.	
Similarly,	given	that	about	64.5%	of	murders	nationally	were	solved	through	arrest	in	
2014,	the	cost	estimates	for	murder	would	likewise	have	to	increased	by	about	55%.		
Instead	of	987	murders,	the	total	by	undocumented	immigrants	would	total	1,529.		
Other	crimes	such	as	rape	have	not	only	low	rates	of	being	reported,	but	only	38.5%	of	
those	that	are	reported	are	solved	through	arrest.17	
Yet,	there	are	two	reasons	even	these	numbers	likely	underestimate	the	amount	of	
crime	committed	by	undocumented	immigrants.	The	most	likely	victims	of	
undocumented	immigrants	are	other	undocumented	immigrants,	and	there	is	a	
common	presumption	frequently	reported	in	the	press	that	undocumented	immigrants	
are	particularly	reticent	to	report	crimes	to	the	police.18		If	undocumented	immigrants	
15

	See	the	FBI	Uniform	Crime	Reports	for	the	clearance	rate	for	each	year.		Clearance	rates	are	the	
percentage	of	crimes	that	are	solved	through	arrest.		For	2014,	the	rates	are	available	here	
(https://ucr.fbi.gov/crime-in-the-u.s/2014/crime-in-the-u.s.-2014/offenses-known-to-lawenforcement/clearances/main).	
	
The	information	on	the	rate	that	crimes	are	reported	to	the	police	is	obtained	by	comparing	the	
estimated	number	of	robberies	in	2014	from	the	National	Crime	Victimization	Survey	with	the	number	of	
crimes	reported	to	police	from	the	Uniform	Crime	Report	
(https://www.bjs.gov/content/pub/pdf/cv14.pdf	and	https://ucr.fbi.gov/crime-in-the-u.s/2014/crime-inthe-u.s.-2014/tables/table-7).	
16
	Give	the	lag	between	arrest	and	conviction,	many	arrests	during	2017	would	not	have	had	time	to	lead	
to	a	plea	agreement	or	a	conviction	and	thus	the	true	conviction	rate	per	arrest	is	actually	higher	than	
51.5%,	though	given	the	data	is	over	6.5	years	this	problem	is	unlikely	too	be	large.		“Texas	Criminal	Alien	
Arrest	Data,”	Texas	Department	of	Public	Safety,	checked	January	3,	2018	
(https://www.dps.texas.gov/administration/crime_records/pages/txCriminalAlienStatistics.htm	and	
https://crimeresearch.org/wp-content/uploads/2018/01/Historical-arrest-and-conviction-data-for-selectoffenses-associated-with-criminal-aliens.jpg).	
17
	It	is	difficult	to	determine	exactly	the	amount	of	underreporting	of	rapes	to	the	police	because	the	
National	Crime	Victimization	Survey	and	the	Uniform	Crime	Reports	don’t	have	exactly	comparable	
numbers.			
18
	Hank	Kalet,	“Prisoners	of	Fear:	NJ’s	Undocumented	Immigrants,”	NJ	Spotlight,	June	16,	2016	
(http://www.njspotlight.com/stories/16/06/15/prisoners-of-fear-nj-s-undocumented-immigrants/);	
Meredith	Hoffman,	“Why	Undocumented	Immigrants	Stay	in	Abusive	Relationships,”	Vice,	March	9,	2016	
(https://www.vice.com/en_us/article/9bg7ma/why-undocumented-immigrants-stay-in-abusiverelationships);	and	Leslie	Berenstein	Rojas,	“Immigrants	a	largely	hidden	segment	of	LA's	homeless	
population,”	KPCC	Public	Radio,	July	14,	2016	
(https://www.scpr.org/news/2016/07/14/62582/immigrants-a-largely-hidden-segment-of-la-s-homele/).		

14
are	indeed	less	likely	to	report	crimes	committed	against	them,	just	using	criminal	
convictions	will	provide	an	underestimate	of	the	true	crime	rate	by	undocumented	
immigrants.		Also,	as	we	will	see,	undocumented	immigrants	tend	to	be	more	involved	
in	gangs	and	those	crimes	are	traditionally	harder	to	solve.	
	
However,	there	are	also	two	possibilities	that	might	work	the	other	way.		One	is	that	
undocumented	immigrants	may	be	relatively	easy	to	catch,	but	that	seems	unlikely.		If	
the	undocumented	immigrant	community	in	an	area	was	very	small,	so	that	their	
members	stood	out	more	readily,	it	would	be	easier	to	catch	them.		But	Arizona	has	an	
unusually	high	rate	of	undocumented	immigrants	with	the	vast	majority	of	them	from	
the	same	country,	Mexico.		In	addition,	as	we	will	see,	these	incarcerations	
overwhelmingly	involve	young	undocumented	immigrants,	who	presumably	are	better	
able	to	adapt	to	a	new	culture	as	they	grow	up	in	the	area.	
	
A	second	consideration	is	that	some	local	authorities	spend	disproportionate	resources	
attempting	to	go	after	undocumented	immigrants.		The	one	clear	example	of	that	would	
be	Sheriff	Joe	Arpaio,	who	was	the	Sheriff	of	Maricopa	County	up	until	the	end	of	2016.		
Maricopa	county	is	the	largest	county	in	the	state	and	makes	up	about	61%	of	Arizona’s	
total	population.	Arpaio	was	famous	for	his	tough	on	illegal	aliens	stand,19	but	on	
January	1,	2017	Democrat	Paul	Penzone	replaced	him.		Yet,	while	we	only	have	data	on	
incarcerations	for	the	first	six	months	of	2017,	the	data	do	not	suggest	that	this	factor	
was	important.		Indeed,	the	share	of	new	prisoners	who	were	undocumented	
immigrants	from	Maricopa	County	increased	from	2016	to	2017,	rising	from	13.26%	to	
14.01%.20		The	level	of	undocumented	immigrants	going	to	prison	from	Maricopa	county	
in	2017	was	greater	than	all	the	years	from	2010	to	2016	and	was	statistically	
significantly	different	from	the	average	for	all	years	at	the	0.00%	level	for	a	two-tailed	ttest.	
	
Thus,	while	incarceration	rates	are	dramatically	higher	for	undocumented	immigrants,	
they	should	be	probably	taken	as	a	lower	bound	when	estimating	how	much	more	likely	
undocumented	individuals	are	convicted	of	crimes.	
	

More	recent	examples	of	such	reporting	includes:	Tom	Dart,	“Fearing	deportation,	undocumented	
immigrants	wary	of	reporting	crime,	“	The	Guardian	(UK),	March	23,	2017	
(https://www.theguardian.com/us-news/2017/mar/23/undocumented-immigrants-wary-report-crimesdeportation)	and	Bryan	Cox,	“Since	Trump,	Latinos	are	reluctant	to	report	crime,”	Newsweek,	May	26,	
2017	(http://www.newsweek.com/trump-latinos-are-reluctant-report-crime-616253).	
19
	See	for	example,	Sarah	Parvini,	“Sheriff	Arpaio	admits	violating	court	order	in	profiling	suit,”	Los	Angeles	
Times,	March	18,	2015	(http://www.latimes.com/nation/la-na-arpaio-immigration-20150318-story.html).	
20
	McCormick	and	Tollison	(1984)	note	that	with	this	type	of	problem	increasing	enforcement	could	either	
increase	or	decrease	the	number	of	arrests,	though	consistent	with	the	interpretation	here,	they	find	
evidence	that	increased	enforcement	reduces	the	number	arrested.	

15
Table	3:	Share	of	Prisoners	for	Arizona	by	Most	Serious	Current	Offense	(Cases	where	
there	are	at	20	incarcerations	for	1985	to	June	2017)	

Offense

Smuggling
Compounding	crime

Undocumented	
immigrants:	
non-US	citizen,	
Not	Legal	
Permanent	
Resident

Number	of	
US	
incarcerations	 Citizens

Percent	increase	
in	the	
Documented	 undocumented	
immigrants:	
immigrants	
non-US	
convicted	
citizen,	Legal	 relative	to	their	
Permanent	 average	share	of	
Resident	
the	Arizona	State	
Population	from	
1985	to	2017

310	

20.6%

78.1%

1.3%

1,526.3%

22	

50.0%

50.0%

0.0%

941.7%

Unlawful	copying	or	sale	of	
sounds	or	images	from	
recording	devices

	
	
27	

Tampering	w/	a	public	
record

63	

	
	

	
	

	
	

51.9%

48.1%

0.0%

903.1%

58.7%

38.1%

3.2%

693.7%

30,208	

66.4%

32.6%

1.0%

579.3%

Criminal	Impersonation

977	

69.1%

30.1%

0.8%

526.9%

Money	Laundering

295	

64.1%

29.2%

6.8%

507.3%

3,160	

72.3%

26.9%

0.8%

459.7%

Marijuana	Violation

Kidnapping
Illegal	control	of	an	
enterprise;	illegally	
conducting	an	enterprise
Theft	by	extortion
Drive	by	shooting;	
forfeiture;	driver	license	
revocation
Discharging	a	firearm	at	a	
structure
Duty	to	give	information	
and	assistance;	alcohol	or	
other	drug	screening
Use	of	wire	
communication	or	
electronic	communication	
in	drug	related	
transactions	

	
	

	
	

	
	

	
	

549	

73.6%

24.8%

1.6%

416.1%

92	

77.2%

22.8%

0.0%

375.5%

78.4%

21.1%

0.4%

339.7%

668	

294	

83	

81.6%

81.9%

	
	
	
	

18.4%

18.1%

	
	
81.4%
204	

	
	
	
	

0.0%

0.0%

	
	
17.6%

	
	
	
	

282.7%

276.5%

	
	
1.0%

	
	
	
	

	
	
267.6%

16
Conducting	a	chop	shop

91	

82.4%

17.6%

0.0%

266.3%

Solicitation

47	

83.0%

17.0%

0.0%

254.6%

Conspiracy

417	

83.5%

16.1%

0.5%

234.7%

Aggravated	Driving/DWI

10,452	

84.1%

15.9%

0.0%

231.1%

Aggravated	DUI

41,243	

81.8%

15.7%

2.5%

227.5%

83.1%

15.6%

1.3%

224.7%

31,949	

85.0%

14.6%

0.4%

204.2%

470	

84.3%

Continuous	sexual	abuse	
of	a	child
Narcotic	Drug	Violation
Discharge	Firearm	in	City	
Limit
Accidents	involving	death	
or	physical	injuries;	failure	
to	stop;	driver	license	
revocation;	restricted	
privilege	to	drive;	alcohol	
or	other	drug	screening

77	

	

	
14.3%

	
	
	
	
	
	

	
1.5%

	
	
	
	
	
	

	
197.0%

	
	
	
	
	
	

	
	
	
	
	
	

562	

84.9%

14.2%

0.9%

196.6%

2,834	

85.2%

13.9%

0.9%

189.6%

59	

86.4%

13.6%

0.0%

182.5%

428	

84.6%

13.6%

1.9%

182.3%

1,706	

85.9%

13.0%

1.1%

171.1%

1,753	

86.1%

12.9%

0.9%

169.8%

nd

2,204	

86.6%

12.8%

0.5%

167.5%

st

1,790	

86.5%

12.6%

0.8%

163.0%

24	

87.5%

12.5%

0.0%

160.4%

4,885	

86.9%

12.1%

1.0%

152.5%

865	

87.9%

12.0%

0.1%

150.5%

2,021	

87.6%

11.7%

0.6%

144.3%

Manslaughter
Facilitation
Participating	in	or	assisting	
a	criminal	syndicate
Sexual	Assault
st

Burglary	in	the	1 	degree
2 	degree	murder
1 	degree	murder
Narcotic	Possess-Transport
Molestation	of	a	child
Possess,	Sell,	Marijuana
Sexual	Abuse
Keeping	or	residing	in	
house	of	prostitution;	
employment	in	
prostitution

	

	
	
	

	

0.0%

	
	
	

	

27	

88.9%

Dangerous	Drug	Violation

32,665	

88.2%

11.1%

0.7%

130.8%

Armed	Robbery

10,493	

88.7%

10.9%

0.4%

127.3%

28	

89.3%

10.7%

0.0%

123.2%

Unsworn	Falsification

11.1%

	
	
	

	

131.5%

	
	
	

17
Negligent	Homicide

757	

88.0%

10.7%

1.3%

122.9%

88.8%

10.6%

0.6%

121.2%

88.7%

10.6%

0.7%

119.9%

Sexual	Conduct	with	a	
Minor

4,597	

Promoting	prison	
contraband;	exceptions;	xradiation;	body	scans

3,250	

Endangerment

6,798	

88.5%

10.4%

1.1%

117.6%

Possession	&	Sale	Narcotic

2,070	

89.6%

10.4%

0.0%

117.4%

88.6%

10.2%

1.1%

113.1%

89.9%

10.1%

0.0%

110.1%

90.0%

10.0%

0.0%

108.3%

89.3%

9.9%

0.8%

106.6%

89.1%

9.9%

1.0%

106.3%

89.2%

9.9%

0.9%

106.0%

88.4%

9.8%

1.7%

104.7%

89.8%

9.4%

0.8%

95.1%

Interference	with	
monitoring	devices
DWI	License	SuspendRevoke

88	
7,525	

DWI	Liquor	or	Drugs

500	

Involving	or	using	minors	
in	drug	offenses

121	

Dangerous	or	deadly	
assault	by	prisoner	or	
juvenile

101	

Taking	identity	of	another	
person	or	entity
DUI	
Liquor/Drugs/VPRS/Combo
Possession	of	burglary	
tools

1,841	
173	
3,503	

Participate	Criminal	Street	

225	

89.3%

9.3%

1.3%

94.4%

Child	Prostitution

129	

89.9%

9.3%

0.8%

93.8%

Escape	1st	Degree

65	

90.8%

9.2%

0.0%

92.3%

38,181	

90.3%

9.1%

0.6%

90.2%

90.6%

9.0%

0.4%

87.4%

91.0%

9.0%

0.0%

86.6%

91.1%

8.9%

0.0%

85.0%

90.7%

8.9%

0.5%

84.5%

Aggravated	Assault		
Unlawful	use	of	means	of	
transportation

8,461	

Obstruction	of	A	Criminal	
Investigation

67	

DWI

1,408	

Sexual	Exploitation	of	a	
Minor

847	

Unlawful	Imprisonment

685	

91.2%

8.8%

0.0%

82.5%

4,202	

90.9%

8.6%

0.5%

79.5%

Flight	from	Law	in	a	

18
Vehicle	
Theft	Means	of	
Transportation

91.1%

8.5%

0.4%

77.3%

95	

89.5%

8.4%

2.1%

75.4%

452	

91.4%

8.4%

0.2%

75.1%

91.1%

8.4%

0.5%

74.1%

3,559	

91.0%

8.2%

0.8%

71.5%

49	

87.8%

8.2%

4.1%

70.1%

141	

92.2%

7.8%

0.0%

62.5%

13,617	

91.9%

7.8%

0.3%

61.6%

Stalking

323	

92.3%

7.7%

0.0%

61.2%

Aggravated	taking	identity	
of	another	person	or	entity

488	

91.4%

Assault	
Criminal	Simulation	
Misconduct	involving	
weapon	
Disorderly	Conduct	
Securing	the	proceeds	of	
an	offense
Prescription-only	drug	
nd

Burglary	2 	Degree

10,425	

10,545	

	

	

	

	

	

	

	

7.6%

	

1.0%

58.0%

	
Table	4:	Share	of	Prisoners	for	Arizona	by	Second	Most	Serious	Current	Offense	(Cases	where	there	are	at	
20	incarcerations	for	1985	to	June	2017)

Undocumented	
immigrants:	
non-US	citizen,	
Not	Legal	
Permanent	
Resident

Percent	
increase	in	the	
undocumented	
Documented	
immigrants	
immigrants:	
convicted	
non-US	
relative	to	their	
citizen,	Legal	
average	share	
Permanent	
of	the	Arizona	
Resident
State	
Population	from	
1985	to	2017

	66	 71.2%

28.8%

0.0%

500%

73.0%

26.0%

1.0%

441%

Facilitation

	3,691	 75.8%

23.1%

1.1%

381%

Solicitation

	15,130	 76.7%

21.7%

1.5%

352%

	208	 78.8%

21.2%

0.0%

341%

	5,523	 79.0%

20.4%

0.6%

324%

	189	 80.4%

19.6%

0.0%

308%

Offense

Marijuana	Violation
Sentence	for	Certain	Drug	
Offenses

Narcotic	Drug	Violation
Conspiracy
Violent	Crimes

Number	of	
incarcerations

	404	

US	
Citizens

19
Attempt
Attempt	to	Commit

	1,490	 82.4%

17.0%

0.5%

255%

	55,533	 83.1%

16.3%

0.6%

240%

82.0%

16.2%

1.8%

238%

85.6%

13.8%

0.6%

188%

	22	 86.4%

13.6%

0.0%

184%

84.2%

13.5%

2.3%

181%

	1,068	 87.1%

12.5%

0.4%

161%

	80	 87.5%

12.5%

0.0%

160%

	220	 85.0%

12.3%

2.7%

156%

	1,034	 87.9%

11.4%

0.7%

138%

89.7%

10.3%

0.0%

114%

90.0%

10.0%

0.0%

108%

	190	 90.0%

10.0%

0.0%

108%

Dangerous	offenders;	
sentencing

	3,807	

Dangerous	crimes	against	
children

	3,410	

Abandonment	of	spouse
Dangerous	crimes	against	
children;	sentences;	
definitions
nd

2 	degree	murder
Assault
Method	of	Inflict	Death
Death	or	Life

	1,284	

Sentence	of	
imprisonment	for	felony

	78	

Notice	of	moving	from	
place	of	residence	or	
change	of	name

	20	

Driving	while	intoxicated
Repetitive	offenders

	12,623	 90.8%

8.5%

0.8%

76%

Danger/Repetitive/Enhan
ce

	37,054	

91.9%

7.8%

0.3%

62%

Sexual	motivation	special	
allegation;	procedures

	66	

87.9%

7.6%

4.5%

58%

Possession	&	Sale	
Narcotic

	27	

92.6%

7.4%

0.0%

54%

Dangerous	Drug	Violation

	141	 92.2%

7.1%

0.7%

48%

DWI	Liquor	or	Drugs

	991	 93.0%

7.0%

0.0%

45%

DWI	License	SuspendRevoke

	156	

94.2%

5.8%

0.0%

20%

93.6%

5.4%

1.0%

13%

91.2%

4.4%

4.4%

-8%

Domestic	violence;	
definition;	weapon	
seizure
DUI	
Liquor/Drugs/VPRS/Comb
o

	4,893	

	68	

20
Offense	Commit	on	
Release

95.6%

4.2%

0.2%

-12%

Shock	Incarceration

	34	 97.1%

2.9%

0.0%

-39%

Soliciting	abortion;	
punishment

	38	

97.4%

2.6%

0.0%

-45%

Aggravated	criminal	
damage

	40	

97.5%

2.5%

0.0%

-48%

Robbery

	68	 98.5%

1.5%

0.0%

-69%

	639	

	
	
Table	5:	Share	of	Prisoners	for	Arizona	by	Most	Serious	Current	Offense	(Cases	where	there	are	at	20	
incarcerations	for	2014)

Offense

Smuggling
Marijuana	Violation

Number	of	
US	
incarcerations Citizens

Undocumented	
immigrants:	
non-US	citizen,	
Not	Legal	
Permanent	
Resident

Documented	
immigrants:	
non-US	
citizen,	Legal	
Permanent	
Resident

Percent	
increase	in	the	
undocumented	
immigrants	
convicted	
relative	to	their	
average	share	
of	the	Arizona	
State	
Population	in	
2014

20	

40%

60%

0%

1150%

1910	

48%

51%

1%

966%

Illegal	control	of	an	
enterprise;	illegally	
conducting	an	enterprise

40	

55%

35%

10%

629%

Money	Laundering

21	

67%

33%

0%

594%

1 	degree	murder

53	

75%

21%

4%

332%

st

97	

81%

15%

3%

222%

201	

82%

15%

3%

211%

st

Burglary	in	the	1 	degree
Molestation	of	Child
Promoting	prison	
contraband;	exceptions;	xradiation;	body	scans;	
classification

	
	

	
	

	
	
	

	
	

	
	
	

	
	

	
	
	

	
	
	

188	

85%

14%

1%

199%

Sexual	Assault

51	

82%

14%

4%

186%

Manslaughter	

104	

83%	

13%	

4%

180%	

Sexual	Abuse

75	

83%

13%

4%

178%

21
Taking	identity	of	another	
person	or	entity

	

	

	

	

148	

86%

13%

1%

167%

Criminal	Impersonation

72	

88%

13%

0%

160%

Sexual	Conduct	with	a	
Minor

220	

87%

12%

1%

146%

Armed	Robbery

466	

87%

12%

1%

141%

Kidnapping

121	

88%

11%

1%

124%

1033	

87%

11%

2%

120%

676	

89%

93	

90%

10%

0%

102%

310	

89%

10%

1%

102%

Aggravated	DUI

1855	

86%

10%

5%

100%

Dangerous	Drug	Violation

2257	

89%

9%

1%

97%

Luring	a	minor	for	sexual	
exploitation

22	

91%

9%

0%

89%

Arson	of	an	occupied	
structure

35	

91%

9%

0%

79%

94	

89%

9%

2%

77%

Flight	from	Law	Vehicle

141	

89%

9%

2%

77%

Aggravated	Robbery

106	

91%

8%

1%

77%

Fraudulent	schemes	and	
artifices

95	

91%

Discharge	Firearm	in	City	
Limit

29	

90%

1852	

93%

Theft	Means	of	
Transportation

441	

94%

5%

0%

13%

Endangerment

250	

91%

5%

4%

8%

nd

606	

94%

5%

1%

7%

rd

655	

95%

5%

0%

5%

21	

95%

5%

0%

-1%

161	

91%

4%

4%

-9%

Narcotic	Drug	Violation
Misconduct	involving	
weapons
Sexual	exploitation	of	a	
minor
Forgery

nd

2 	degree	murder

Aggravated	Assault

	

	
10%

	

	

	

7%
6%

	

Burglary	in	the	2 	degree
Burglary	in	the	3 	degree	
Negligent	Homicide
Organized	retail	theft

	

	
1%

	

	

	

	
8%

	
106%

	

	

	
	

	
1%

3%

	

1%
	

	
75%
44%

	

25%
	

	

22
Arson	of	
Structure/Property

24	

nd

Escape	2 	Degree

48	

Trafficking	in	stolen	
property

364	

92%

4%

4%

-13%

96%

4%

0%

-13%

95%

4%

1%

-14%

4%

0%

-15%

Unlawful	Imprisonment

49	

96%

Aggravated	taking	identity	
of	another	person	or	entity

51	

94%

4%

2%

-18%

Child/Adult	Abuse

154	

94%

4%

2%

-19%

Theft

592	

95%

4%

1%

-23%

Robbery

163	

96%

4%

0%

-23%

Resisting	Arrest

200	

97%

4%

0%

-27%

Shoplifting

283	

96%

3%

0%

-34%

Failure	to	Register	as	a	sex	
offender

129	

97%

3%

0%

-35%

Disorderly	Conduct

194	

95%

3%

2%

-36%

37	

97%

3%

0%

-44%

96%

3%

2%

-46%

97%

3%

0%

-47%

96%

3%

1%

-47%

97%

2%

0%

-51%

96%

2%

1%

-54%

98%

2%

0%

-56%

98%

2%

0%

-62%

98%

2%

1%

-66%

97%

2%

1%

-67%

140	

99%

1%

0%

-85%

29	

97%

0%

3%

-100%

Threat-Intimidate
Criminal	Possession	
Forgery	Document

	

	

115	

Sex	Offender	Registry	
Violation

78	

Unlawful	use	of	means	of	
transportation

277	

Aggravated	domestic	
violence

211	

Criminal	Damage

136	

Possession	of	burglary	
tools

190	

Aggravated	Harassment

55	

Drug	Paraphernalia	
Violation
Criminal	trespass	in	the	1
degree
Theft	Credit	Card
Failure	to	appear	in	the	

	

1723	

	

	

	

	

	

st	

317	

23
first	degree	
Notice	of	moving	from	
place	of	residence	or	
change	of	name	or	
electronic	information;	
forwarding	of	information

	
	
	
	
30	

100%

	
	
	
	
0%

	
	
	
	
0%

	
	
	
	
-100%

Do	changes	in	enforcement	along	the	Mexican-U.S.	Border	affect	the	incarceration	
rate	of	Undocumented	Immigrants?	
One	concern	with	the	preceding	results	is	that	they	are	being	driven	by	how	porous	the	
Arizona	border	is	with	Mexico.		The	concern	is	that	as	illegals	pass	through	Arizona	on	
their	way	to	other	jurisdictions,	they	will	commit	crime.		If	the	census	estimates	pick	up	
those	who	are	temporarily	in	Arizona,	this	effect	shouldn’t	matter	since	even	though	
there	might	be	a	different	person	in	Arizona	in	June	than	in	December,	the	total	in	the	
denominator	is	all	that	matters.	Yet,	it	may	be	more	difficult	to	measure	those	who	are	
only	in	Arizona	for	short	periods	of	time.	The	small	confidence	intervals	claimed	by	
those	the	Census	data	belie	that	they	have	considered	this	a	serious	problem.	
If	this	concern	is	important,	the	size	of	the	effect	should	vary	over	time	as	border	
enforcement	between	Arizona	and	Mexico	changes.		A	more	porous	should	be	related	
to	a	higher	share	of	undocumented	immigrants	among	those	who	are	newly	
incarcerated.			
There	were	two	major	changes	in	enforcement	during	the	period	that	we	study.		In	
1994,	the	Clinton	Administration	started	“Operation	Gatekeeper,”	which	“succeeded	in	
shutting	off	many	of	the	California	routes”	that	undocumented	immigrants	took	into	the	
U.S.21		This	change	appeared	to	occur	over	time	between	1994	and	2000.		Cutting	off	
this	route	into	California	was	associated	with	a	surge	in	undocumented	immigrants	into	
Arizona.		In	response	in	2005,	border	enforcement	was	then	beefed	up	in	Arizona,	and	
undocumented	immigrants	began	to	enter	the	U.S.	through	Texas.	
If	this	“porous	border”	hypothesis	is	correct,	these	changes	in	enforcement	imply	that	
undocumented	immigrants	share	of	newly	incarcerated	individuals	should	increase	
between	1994	and	2000	and	then	fall	after	2005.		In	fact,	nothing	like	that	seems	to	
have	occurred.		Between	1994	and	2000,	the	percent	of	those	incarcerated	each	year	
who	were	undocumented	immigrants	fell	from	12.2%	in	1994	to	9.7%.		Even	by	2004,	it	
was	almost	the	same	level	that	it	was	in	1994	(12.0%	versus	12.2%).		Similarly,	after	
2005,	instead	of	this	percent	declining	it	initially	rose,	going	from	11.9%	in	2005	to	
13.8%	in	2009.		After	that,	the	percentage	fell	and	then	rose	again,	but	none	of	the	
21

	Edward	Alden,	“Arizona’s	Alarm	Bell	for	Immigration	Reform,”	Council	on	Foreign	Relations,	April	26,	
2010	(https://www.cfr.org/expert-brief/arizonas-alarm-bell-immigration-reform).		See	also	Roberts	(2017)	
for	evidence	of	the	impact	of	border	enforcement	on	rate	that	undocumented	immigrants	.	

24
variation	seems	capable	of	explaining	the	large	difference	in	incarceration	rates	
between	undocumented	immigrants	and	U.S.	citizens.	
Nor	is	Arizona	a	particularly	important	route	for	drug	trafficking.22		According	to	the	
Department	of	Homeland	Security,	cocaine,	heroin,	and	meth	enter	through	ports	of	
entry,	mainly	by	being	concealed	in	cars	and	trucks.	This	means	that	major	entry	
corridors	for	these	drugs	are	in	southern	California	and	Texas,	because	that's	where	
most	of	the	vehicle	entry	takes	place.		A	significant	amount	of	marijuana	comes	through	
the	Border	States,	with	the	exception	of	New	Mexico.	
	
One	change	left	unexplained	is	the	large	run	up	in	the	percentage	of	newly	incarcerated	
people	who	were	undocumented	immigrants	between	1985	and	1994.		There	are	not	
enough	years	of	data	here	to	determine	if	this	was	a	result	1986	amnesty	that	was	
granted	by	President	Reagan.		If	the	share	of	undocumented	immigrants	who	are	
prisoners	is	a	proxy	for	their	share	of	the	population,	it	is	possible	that	the	amnesty	
increased	undocumented	immigrants	coming	into	the	U.S.		Alternatively,	the	amnesty	
might	be	related	to	the	type	of	undocumented	immigrant	coming	to	the	US	or	their	
ability	to	commit	crime.		More	work	remains	on	these	questions.	

	
22

	“Drug	Smuggling	at	the	Border,”	U.S.	Department	of	Homeland	Security,	October	18,	2017.	

25
	
	
Recidivism	
	
Some	of	the	other	characteristics	of	undocumented	immigrant	criminals	stand	out	and	
we	will	go	through	the	implications	of	this	in	the	next	sections.		Undocumented	
immigrant	criminals	are	45.4%	more	likely	to	have	been	gang	members,	they	are	133%	
more	likely	to	receive	sentencing	enhancements	for	being	classified	as	dangerous,	and	
they	tend	to	be	released	from	prison	at	a	younger	age	despite	more	serious	crimes	
because	they	initially	go	to	jail	at	a	younger	age.23		These	factors	are	clearly	related	to	
recidivism,	but	affect	it	in	different	directions.		Higher	gang	membership	and	lower	ages	
for	release	from	prison	normally	tend	to	mean	higher	recidivism	rates.		Sentencing	
enhancements	for	more	violent	behavior	is	actually	associated	with	a	lower	rate	of	
recidivism.		
	
While	convictions	of	undocumented	immigrants	are	high	rate	relative	to	their	share	of	
the	population,	their	recidivism	rates	are	extremely	low.	For	example,	only	2.3	of	
convicted	undocumented	immigrants	who	are	considered	dangerous	and	who	have	a	
history	of	repetitive	behavior	recidivate	within	6	years	of	being	released.		In	contrast,	
convicted	U.S.	Citizens	with	characteristics	exhibit	a	six-year	recidivism	rate	of	26.4%	–		
about	9.7	times	higher.			
	
Anyway	you	cut	the	data,	U.S.	citizens	recidivate	at	much	higher	rates.		For	example,	
when	compared	with	undocumented	immigrants	with	equivalent	sentencing	
enhancements	or	gang	affiliations,	U.S.	Citizens	are	2	to	11.5	times	more	likely	to	
recidivate	within	six	years	than	are	undocumented	immigrants.		For	ages	under	65,	U.S.	
citizens	are	3.6	to	4.7	times	more	likely	to	recidivate	within	six	years.	
	

23

	The	gang	membership,	sentencing	enhancements,	and	other	values	discussed	in	the	text	are	simple	
conditional	means.		This	general	pattern	is	maintained	using	logit	regressions	even	when	other	factors	are	
accounted	for	(Appendix	4).		Hispanic	undocumented	immigrants	are	more	likely	to	be	gang	members,	
face	longer	prison	sentences,	are	younger,	more	likely	to	be	male,	and	less	likely	to	use	a	weapon	or	to	
injure	someone.		Hispanic	U.S.	citizens	are	even	more	likely	to	be	gang	members	than	Hispanics	who	are	
not	U.S.	citizens,	and	they	are	also	more	likely	to	use	weapons	and	injure	someone	but	less	likely	to	be	
male.			Native	Americans	are	the	only	other	groups	that	are	more	likely	to	gang	members.	
	
Some	of	the	results	are	quite	large.		Being	suspected	of	gang	membership	lowers	the	odds	of	the	prisoner	
being	white	by	75%	and	black	by	over	95%,	but	it	raises	the	odds	of	the	person	being	an	undocumented	
Hispanic	immigrant	by	32%.		If	the	prisoner	was	convicted	of	a	weapons	charge	involving	a	gun,	it	lowers	
the	odds	of	the	prisoner	being	white	by	34%	and	Native	American	by	45%,	but	raises	the	odds	of	the	
prisoner	being	black	by	59%.		Incarcerated	undocumented	Hispanic	immigrants	are	also	almost	exclusively	
male	to	a	much	higher	rate	than	any	other	type	of	prisoner.		
	

26
	

Table	6:	Recidivism	Rates	by	Miscellaneous	Risk	Factors	for	Arizona	by	U.S.	Citizen	
and	Undocumented	Immigrants
US	Citizens

Undocumented	immigrants

Number	of	Years/Percent	Recidivated

Number	of	Years/Percent	Recidivated

Gang	
Affiliation	
Status

1

2

3

4

5

6

1

2

3

4

5

6

21.2	

27.8	

31.1	

31.8	

33.8	

Debriefed

36.9	

52.1	 60.2	 64.3	 66.2	 67.8	 13.9	

Participant

12.5	

12.5	 25.0	 37.5	 37.5	 37.5	 	

Step	Down

29.8	

47.2	 54.0	 57.8	 60.9	 62.7	

0.0	

0.0	

0.0	

16.7	

16.7	

16.7	

Suspect

32.4	

45.8	 53.4	 57.7	 60.3	 62.1	

5.1	

9.3	

11.7	

13.8	

14.6	

15.6	

Validated

33.8	

47.2	 54.8	 58.3	 60.9	 61.9	

6.9	

13.8	

17.4	

18.6	

21.0	

22.2	

Missing

30.0	

40.0	 60.0	 60.0	 60.0	 70.0	

	

	

	

	

	

	

Total

32.5	 						46.0	 53.6	 57.9	 60.5	 62.2	

5.5	

9.9	

12.4	

14.5	

15.5	

16.5	

	

	

	

	

	

	

	

	

	

	

	

	

	

Number	of	Years/Percent	Recidivated	
History	of	
Sentence	
Enhanceme
nt

1	

2	

3	

4	

5	

	

	

	

	

	

Number	of	Years/Percent	Recidivated	

6	

1	

2	

3	

4	

5	

6	

NOT	
Dangerous	
(ND)/	NOT	
REPETITIVE	
(NR)
21.8	

30.6	 36.1	 39.5	 41.8	 43.4	

3.0	

5.3	

7.2	

8.5	

9.5	

10.1	

Dangerous	
(D)/NR

11.5	

16.3	 19.0	 20.4	 21.6	 22.2	

0.9	

1.4	

2.1	

2.5	

2.7	

2.8	

Dangerous	
(D)/	
Repetitive	
(R)

14.9	

19.6	 22.4	 24.0	 25.6	 26.4	

1.2	

1.2	

2.3	

2.3	

2.3	

2.3	

ND/R

24.5	

33.8	 38.7	 41.6	 43.5	 44.7	

3.1	

5.6	

7.1	

8.1	

8.9	

9.5	

N/A

11.2	

17.0	 21.1	 23.5	 24.8	 25.9	

3.4	

4.5	

5.6	

5.6	

5.6	

5.6	

Total

21.6	

30.3	 35.6	 38.9	 41.1	 42.7	

2.9	

5.1	

6.9	

8.2	

9.0	

9.7	

27
	

	

	

	

	

	

	

Number	of	Years/Percent	Recidivated	
Age	at	
Release

2	

3	

0-18

32.1	 46.7	

19

	

	

Number	of	Years/Percent	Recidivated	
2	

3	

4	

5	

6	

56.5	

61.3	 64.2	 66.2	

7.5	

13.3	

15.0	

16.8	

18.6	

18.6	

34.4	 49.6	

57.7	

62.0	 64.6	 66.5	

6.3	

8.9	

11.7	

13.1	

14.3	

15.2	

20-21

29.6	 42.9	

50.4	

54.8	 57.7	 59.7	

4.3	

7.5	

10.3	

12.3	

13.4	

14.4	

22-24

26.5	 38.1	

45.2	

49.4	 52.4	 54.4	

3.6	

6.6	

8.8	

10.6	

11.9	

12.6	

25-29

24.5	 35.3	

41.8	

45.8	 48.5	 50.5	

3.3	

6.3	

8.5	

10.2	

11.2	

11.9	

30-34

24.5	 34.7	

40.8	

44.6	 47.1	 49.0	

3.2	

5.8	

7.8	

9.3	

10.3	

11.0	

35-39

24.1	 33.1	

39.0	

42.7	 45.2	 46.9	

3.5	

5.8	

7.8	

9.1	

10.2	

11.0	

40-44

22.9	 31.2	

36.6	

40.1	 42.6	 44.3	

2.8	

4.9	

6.8	

8.0	

8.8	

9.5	

45-49

21.2	 28.5	

33.0	

36.1	 37.9	 39.2	

2.9	

4.8	

6.7	

7.8	

8.6	

9.2	

50-54

18.6	 24.8	

28.3	

30.5	 32.1	 33.1	

2.7	

4.6	

6.0	

6.9	

7.8	

7.9	

55-59

16.3	 21.0	

23.9	

25.6	 26.8	 27.5	

3.0	

4.3	

5.2	

5.7	

6.6	

7.6	

60-64

12.5	 15.3	

17.4	

18.4	 19.3	 19.8	

2.1	

3.3	

3.8	

5.0	

5.4	

5.4	

65-69

10.3	 13.1	

14.3	

15.1	 15.3	 15.6	

3.8	

4.7	

6.6	

8.5	

8.5	

8.5	

9.3	

4.2	

5.6	

7.0	

8.5	

9.9	

9.9	

42.6	 45.0	 46.7	

3.4	

5.9	

8.0	

9.5	

10.5	

11.2	

Total

7.4	

8.2	

23.7	 33.2	

39.0	

8.8	

5	

	

1	

6.2	

4	

	

6	

70+

1	

	

9.1	

	
	

The	vast	differences	in	prison	re-entry	rates	among	U.S.	citizens,	undocumented	
immigrants,	and	documented	immigrants	is	clear	by	looking	at	their	differing	criminal	
conviction	histories.		Undocumented	immigrants	have	a	very	high	share	among	those	
who	have	been	convicted	and	sentenced	to	prison	once	or	twice,	but	U.S.	citizens	are	
vastly	more	likely	to	have	had	at	least	three	admissions	into	the	Arizona	Department	of	
Corrections	system.	

28
	
	
Table	7:	Criminal	History	for	U.S.	citizens,	undocumented	immigrants,	and	
documented	immigrants:	Arizona	(Each	row	sums	to	100%)

US	Citizens

Undocumented	
immigrants:	nonUS	citizen,	Not	
Legal	Permanent	
Resident

Documented	
immigrants:	nonUS	citizen,	Legal	
Permanent	
Resident

	
Ratio	of	U.S.	
Citizens	relative	to	
Undocumented	
immigrants

1

76.5%

22.3%

1.3%

3.43

2

88.9%

10.4%

0.7%

8.55

3

94.3%

5.2%

0.5%

17.98

4

96.8%

3.0%

0.2%

31.84

5

97.6%

2.3%

0.1%

42.55

6

98.4%

1.5%

0.1%

63.74

7

98.7%

1.3%

0.0%

76.39

8

99.1%

0.9%

0.0%

116.55

9

99.0%

1.0%

0.0%

97.67

10

99.5%

0.5%

0.0%

188.68

11

99.3%

0.7%

0.0%

139.69

12

99.3%

0.7%

0.0%

144.38

13

99.2%

0.8%

0.0%

129.00

14

100.0%

0.0%

0.0%

15

96.4%

1.4%

2.1%

16

100.0%

0.0%

0.0%

17

97.6%

2.4%

0.0%

41.00

Total

88.1%

11.2%

0.7%

7.88

Number	of	
admissions	into	
Arizona	
Department	of	
Corrections

67.50

	
	
A	significant	number	of	U.S.	citizens	just	keep	cycling	through	the	Arizona	prison	system.		
24.76%	of	U.S.	citizens	convicted	in	Arizona	had	served	in	the	Arizona	Department	of	
Corrections	at	least	five	times.		That	simply	isn’t	the	case	for	undocumented	immigrants,	
where	only	2.95%	of	them	had	the	same	prison	history.			

29

	
	
Table	8:	Criminal	History	for	U.S.	citizens,	undocumented	immigrants,	and	
documented	immigrants:	Arizona

US	Citizens

Undocumented	
immigrants:	nonUS	citizen,	Not	
Legal	Permanent	
Resident

Documented	
immigrants:	nonUS	citizen,	Legal	
Permanent	
Resident

	
Ratio	of	U.S.	
Citizens	relative	to	
Undocumented	
immigrants

5	or	more	times

24.76%

2.95%

0.08%

8.40

6	or	more	times

16.44%

1.53%

0.04%

10.75

7	or	more	times

10.54%

0.78%

0.02%

13.45

8	or	more	times

6.65%

0.36%

0.02%

18.30

9	or	more	times

3.97%

0.20%

0.01%

19.80

10	or	more	times

2.38%

0.08%

0.01%

30.38

Number	of	
admissions	into	
Arizona	
Department	of	
Corrections

	
	
It	is	remarkable	that	undocumented	immigrants	make	up	such	a	large	share	of	the	
prison	population	given	that	they	show	such	a	low	recidivism	rate	and	have	a	relatively	
low	number	of	admissions	to	prison.		Unlike	U.S.	citizens,	thus	it	is	rarely	the	same	
undocumented	immigrants	going	in	and	out	of	prison.		Thus	a	much	larger	share	of	
undocumented	immigrants	are	committing	crime	compared	to	U.S.	citizens	than	would	
normally	be	inferred	from	just	looking	at	undocumented	immigrants	share	of	those	
entering	prison.	
	
While	undocumented	immigrants	are	convicted	of	more	serious	crimes	and	face	longer	
prison	terms,	that	is	more	than	offset	by	the	younger	ages	at	which	undocumented	
immigrants	tend	to	be	convicted.		Nor	does	it	seem	likely	that	the	different	criminal	
histories	result	from	undocumented	immigrants	being	more	difficult	to	catch	or	having	a	
greater	tendency	to	reform	after	prison.		After	all,	young	undocumented	immigrants	
have	a	high	incarceration	rate,	so	it	isn't	obvious	why	their	incarceration	rate	for	any	
further	crimes	should	then	fall	so	much	relative	to	U.S.	citizens	once	they	have	served	
time	in	prison.			
	
As	to	the	alternative	that	the	undocumented	immigrants	are	reformed	after	their	first	or	
second	time	in	prison,	the	low	rate	of	recidivism	seems	impossibly	low	and	has	no	
parallels	for	any	other	groups.		How	can	dangerous	convicts	with	repetitive	behavior	
have	a	recidivism	rate	of	about	two	percentage	points?		Nor	does	it	seem	likely	that	

30

gang-affiliated	undocumented	immigrants	would	have	a	recidivism	rate	that	is	just	a	
fourth	of	the	rate	of	U.S.	citizens	who	are	in	gangs.	
	
Further	evidence	on	this	last	point	comes	from	Mexican	politicians	who	complain	about	
the	crime	committed	by	those	criminals	deported	back	to	their	country	after	they	have	
served	their	time	in	prison.24			
	
The	most	likely	explanation	for	the	low	recidivism	rate	is	that	many	undocumented	
immigrants	don’t	return	to	Arizona	or	the	U.S.	when	they	are	deported	after	being	
released	from	prison.	
	
This	also	has	additional	implications	for	how	hard	it	is	to	catch	undocumented	
immigrants	who	commit	crime.		It	is	more	difficult	for	police	to	identify	criminals	who	
have	no	criminal	record.		For	example,	people	who	have	previously	been	convicted	of	
crime	have	their	DNA	and	fingerprints	in	law	enforcement	databases.		This	then	
provides	yet	another	reason	why	the	numbers	presented	here	likely	underestimate	the	
amount	of	crime	committed	by	undocumented	immigrants.	
	
The	data	also	provides	a	unique	view	into	recidivism	rates	when	other	factors	are	
accounted	for.		A	simple	Logit	regression	examines	the	recidivism	rate	by	the	convict’s	
sixth	year	after	release	(Table	9).		The	estimates	account	for	citizenship,	gender,	race,	
whether	the	convict	committed	a	crime	that	injured	someone,	the	weapon	used,	
suspected	gang	membership,	the	age	at	release	from	prison,	the	number	of	days	
incarcerated,	and	fixed	effects	for	the	year	of	release	and	jurisdiction	that	the	inmate	
came	from.		The	estimates	are	interpreted	as	the	odds	of	recidivism	relative	to	that	for	
white	female	U.S.	citizen	who	has	no	weapon,	wasn’t	a	gang	member,	and	who	didn’t	
injure	anyone.			
	
The	most	striking	findings	are	that	illegal	aliens,	older	releasees,	and	people	serving	
longer	sentences	even	when	the	age	of	release	is	accounted	for	have	lower	recidivism	
rates.		A	ten-year	increase	in	age	lowers	the	odds	of	recidivism	by	12%.		Gang	members,	
Hispanics	who	are	born	in	the	US,	and	Native	Americans	have	much	higher	recidivism	
rates.	For	example,	the	odds	of	someone	suspected	of	being	a	gang	member	
recidivating	are	169%	higher.	
	
The	estimates	imply	that	the	odds	of	an	undocumented	immigrant	recidivating	are	30%	
lower	than	for	a	U.S.	citizen.		If	the	person	is	an	undocumented	Hispanic	immigrant,	the	
odds	of	them	recidivating	are	over	50%	lower.	
24

	Diane	Macedo,	“U.S.	Worsens	Mexican	Violence	by	Returning	Criminal	Aliens	to	Border	Cities,	Mayors	
Say,”	Fox	News,	September	29,	2010	(http://www.foxnews.com/us/2010/09/28/mexican-lawmakers-sayworsening-mexico-violence-returning-criminals.html).	

31

	
An	F-test	for	the	first	estimate	shows	that	Hispanics	who	were	born	in	the	US	have	a	
much	higher	recidivism	rate	than	those	who	aren’t	born	in	the	US	is	significant	at	the	
0.0000	level.		It	is	clear	that	the	key	factor	here	isn’t	whether	someone	is	Hispanic,	but	
whether	they	are	from	the	US	or	not.		While	the	odds	of	recidivism	are	25%	higher	for	
Hispanics	born	in	the	US,	the	odds	are	31%	lower	for	those	born	in	another	country.		
This	provides	evidence	that	those	deported	are	much	less	likely		
	
	
Table	9:		Explaining	Differences	in	Recidivism	Rates:	Logit	regression	to	see	if	convict	
recidivates	by	the	sixth	year	after	they	are	released	from	prison		
(Also	accounts	for	year	and	jurisdiction	fixed	effects	and	felony	class	type)	
Variable	
Coefficient	
Absolute	
Z-statistic	
Undocumented	
-.364	
2.51**	
Immigrant	
Documented	
-.157	
0.42	
Immigrant	
Male	
-.059	
1.40	
Black	
.0497	
1.18	
Native	American	
.186	
3.42***	
Hispanic	Born	US	
.2197	
6.52***	
Hispanic	Not	Born	
-.368	
2.43**	
US	
Asian	
-.194	
0.54	
Race	Other	
-.075	
0.42	
Injury		
.0456	
0.66	
Weapon	-	Gun	
-.031	
0.43	
Weapon	-	Knife	
-.131	
1.17	
Weapon	-	Other	
-.069	
0.85	
Weapon	-	
.0183	
0.38	
Unknown	
Weapon	-	vehicle	
-.0034	
0.04	
Suspected	Gang	
0.988	
2.73***	
member	
Age	at	release	
-.0123	
7.83***	
Number	of	days	
-.00013	
5.04***	
incarcerated	
Observations	
334,629	
Log	likelihood	
-27968.325	
Chi	Square	
1768.51d	
*	Statistically	significant	at	the	10	percent	level	for	a	two-tailed	z-test.	
**	Statistically	significant	at	the	5	percent	level	for	a	two-tailed	z-test.	
***	Statistically	significant	at	the	1	percent	level	for	a	two-tailed	z-test.	

32
The	Age	Distribution	of	Incarcerated	Criminals	by	Citizenship	Status	
	
While	we	know	that	undocumented	immigrants	are	disproportionately	likely	to	be	
convicted	of	crimes	than	citizens	or	documented	immigrants,	one	question	is	how	that	
varies	with	age.		The	issue	of	age	is	particularly	relevant	given	the	debate	over	Deferred	
Action	for	Childhood	Arrivals	(DACA),	which	raises	questions	about	the	age	distribution	
of	convicts.		To	be	eligible	for	DACA,	undocumented	immigrants	have	to	be	born	after	
June	15,	1981	and	entered	into	the	U.S.	since	June	15,	2007.25		By	the	end	of	June	2017,	
the	eligible	ages	for	undocumented	immigrants	was	between	the	age	of	15	and	those	
who	were	two	weeks	past	their	36th	birthday.		DACA	also	has	education	and	criminal	
history	requirements,	though	we	don’t	have	information	on	educational	background	for	
those	who	are	prisoners.		As	to	criminal	history,	the	fact	undocumented	immigrants	
tend	to	commit	just	one	or	two	crimes	means	that	criminal	histories	are	much	less	
useful	in	distinguishing	undocumented	immigrants	than	the	rest	of	the	population.	
	
The	Migration	Policy	Institute	provides	an	age	distribution	of	the	undocumented	
population	for	the	U.S.	that	indicates	that	47%	are	from	15	through	35	years	of	age	and	
45.7%	are	36	and	older.26,27	In	contrast,	the	same	percentages	for	the	Arizona	legal	
populations	are	28.1%	and	49.9%	respectively.		Some	of	the	gap	for	the	15	to	35	age	
group	arises	from	there	being	very	few	undocumented	immigrants	under	the	age	of	15	
(only	7.3%	of	all	undocumented	immigrants).	This	occurs	because	many	undocumented	
immigrants	had	children	who	were	born	in	the	United	States.	
	
25

	U.S.	Citizenship	and	Immigration	Services,	Department	of	Homeland	Security,	“Consideration	of	
Deferred	Action	for	Childhood	Arrivals	(DACA),”	https://www.uscis.gov/archive/consideration-deferredaction-childhood-arrivals-daca.		See	also	Migration	Policy	Institute,	“Deferred	Action	for	Childhood	
Arrivals	(DACA)	Data	Tools,”	Data	Hub,	http://www.migrationpolicy.org/programs/data-hub/deferredaction-childhood-arrivals-daca-profiles.	
26
	The	age	ranges	provided	by	the	Migration	Policy	Institute	don’t	quite	match	what	we	are	looking	for	so	
we	assumed	that	the	number	of	people	per	year	were	constant	within	their	age	group	ranges.	“Profile	of	
the	Unauthorized	Population:	United	States,”	Migration	Policy	Institute	
(https://www.migrationpolicy.org/data/unauthorized-immigrant-population/state/US).	
27
	Another	estimate	from	the	Migration	Policy	Institute	implies	a	lower	percent	of	the	population	who	are	
in	the	15	to	35	year	old	age	group	and	it	implies	a	higher	crime	rate	for	them.		It	puts	the	number	of	
undocumented	immigrants	in	this	range	at	1.9	million	who	arrived	in	the	U.S.	prior	to	2008.	That	would	
imply	they	represent	just	16.8%	of	the	undocumented	immigrants	in	the	U.S..	If	these	young	
undocumented	immigrants	entered	into	the	U.S.	at	the	same	proportion	over	time,	another	4	percentage	
points	would	be	added	to	the	total.		Assuming	that	percentage	holds	true	for	Arizona,	1.01%	percent	of	
the	total	population	are	undocumented	immigrants	from	15	to	35	years	old.	This	number	includes	those	
who	have	not	met	the	education	requirement.		In	addition,	“ineligibility	due	to	criminal	history	or	lack	of	
continuous	U.S.	presence	were	not	modeled	due	to	lack	of	data.	MPI	estimates	of	the	DACA-eligible	
population”	(p.	11).		Randy	Capps,	Michael	Fix,	and	Jie	Zong,	“The	Education	and	Work	Pro	les	of	the	DACA	
Population,”	Migration	Policy	Institute,	August	2017	(available	here	
https://www.migrationpolicy.org/research/education-and-work-profiles-daca-population).		“Deferred	
Action	for	Childhood	Arrivals	(DACA)	Data	Tools,”	Data	Hub,	
http://www.migrationpolicy.org/programs/data-hub/deferred-action-childhood-arrivals-daca-profiles.	

33
We	calculate	shares	of	the	prison	population	based	on	the	age	at	which	the	criminal	
entered	prison.		So	undocumented	immigrants	between	15	and	35	make	up	2.27%	of	
the	total	population	and	7.94%	of	convicts.		While	the	legal	population	between	15	and	
35	represents	26.7%	of	the	total	population,	they	account	for	just	54.7%	of	the	legal	
population	in	prison.28		Young	undocumented	immigrants	make	up	a	71%	greater	share	
of	their	group’s	share	of	the	prison	population	relative	to	their	group’s	share	of	the	
general	population	than	the	same	ratio	for	legal	residents.	
	
A	similar	calculation	can	be	made	for	those	over	35.		Such	undocumented	immigrants	
account	for	3.22%	of	the	prison	population	and	2.2%	of	the	general	population.		In	
comparison,	the	legal	population	accounts	for	33.8%	of	prisoners	and	47.4%	of	the	
state’s	total	population.		Adjusting	for	the	lower	crime	rate	that	older	people	commit,	
older	undocumented	immigrants	are	about	104%	more	likely	to	be	convicted	than	the	
rest	of	the	older	population.	
Table	10:	Comparing	share	of	convicted	criminals	entering	prison	by	year	of	age	and	
citizenship	status	for	Arizona	(1985	to	June	2017)	
	

Age	at	admission	
for	prison	
15-20	
21-25	
26-30	
31-35	
36-40	
41-45	
46-50	
51-55	
56-60	
61-65	
66-70	
71+	
Total	percent	by	
citizenship	status	
for	those	under	36	
Total	percent	by	
citizenship	status	
for	those	at	least	
36	years	of	age	

Share	of	total	people	entering	prison	by	
citizenship	status	
Undocumented	
immigrants:	non-US	
citizen,	Not	Legal	
Permanent	
US	Citizens	
Resident	
5.49%	
0.86%	
16.86%	
2.45%	
16.99%	
2.57%	
15.21%	
2.10%	
12.42%	
1.47%	
9.33%	
0.87%	
6.06%	
0.48%	
3.32%	
0.24%	
1.48%	
0.10%	
0.60%	
0.04%	
0.23%	
0.02%	
0.12%	
0.01%	

54.6%	

7.97%	

33.6%	

3.22%	

	
Undocumented	immigrants	share	of	
those	entering	prison	as	a	percent	of	
their	share	of	the	population	(2.27%	
of	population	for	those	15	to	35	years	
old	and	2.20%	of	those	36+)	
	

	
	
251%	
	
	
	
46.3%	

	
28

	US	Census	Bureau,	American	Fact	Finder,	Age	and	Sex,	
https://factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?src=CF	

34
Unfortunately,	if	the	goal	of	DACA	is	to	give	citizenship	to	a	particularly	law-abiding	
group	of	undocumented	immigrants,	it	is	accomplishing	the	opposite	of	what	was	
intended.		As	Table	10	shows,	DACA	age	eligible	undocumented	immigrants	are	250%	
more	likely	to	be	convicted	of	crimes	than	their	share	of	the	population.		Those	too	old	
for	DACA	status	are	convicted	at	a	relatively	low	rates	(45.7%	more	than	their	share	of	
the	Arizona	population).			
	
To	put	it	differently,	the	actual	number	of	young	undocumented	immigrants	in	the	
Arizona	would	have	to	be	almost	nine	times	greater	than	estimated	number	that	
researchers	have	provided	before	their	rate	of	incarceration	would	be	the	same	as	the	
general	population.			
	
The	DACA	age	eligible	undocumented	immigrants	also	commit	more	serious	crimes	and	
serve	sentences	that	are	about	9.7%	longer	than	those	for	U.S.	citizens.	
	
As	noted	previously,	given	the	low	rate	that	crimes	are	generally	reported	and	solved,	
many	criminals	would	be	eligible	for	citizenship	under	DACA	and	the	high	share	that	
undocumented	immigrants	make	of	those	from	15	to	35	years	of	age	who	are	
imprisoned	raises	real	concerns	about	the	over	all	crime	rate	for	young	undocumented	
immigrants	who	do	not	have	a	criminal	record.			If	undocumented	immigrants	are	even	
less	likely	than	average	to	be	caught	because	their	victims	are	reticent	to	report	crimes,	
this	discussion	underestimates	the	problem	with	DACA	age	eligible	undocumented	
immigrants.

35
	
	
Table	11:	Share	of	Prisoners	from	15	through	35	years	of	age	at	age	of	admission	for	
Arizona	by	Most	Serious	Current	Offense	(Cases	where	there	are	at	20	incarcerations	for	
1985	to	June	2017)

Undocumented	
immigrants:	
non-US	citizen,	
Not	Legal	
Permanent	
Resident

Documented	
immigrants:	
non-US	
citizen,	Legal	
Permanent	
Resident

Percent	
increase	in	the	
undocumented	
immigrants	
convicted	
relative	to	their	
average	share	
of	the	Arizona	
State	
Population	
(assuming	that	
rate	is	2.27%)

15.9%	

84.1%	

0.0%	

3604.9%	

55.6%	

42.2%	

2.2%	

1760.0%	

20,228	

64.0%	

35.2%	

0.8%	

1452.3%	

Criminal	Impersonation	

656	

69.1%	

30.6%	

0.3%	

1249.8%	

Money	Laundering	

152	

64.5%	

30.3%	

5.3%	

1233.2%	

2,260	

69.8%	

29.6%	

0.6%	

1202.1%	

	
	
303	

70.3%	

28.7%	

1.0%	

1164.9%	

Theft	by	extortion	

63	

73.0%	

27.0%	

0.0%	

1088.7%	

Conducting	a	chop	shop	

48	

77.1%	

22.9%	

0.0%	

909.5%	

Drive	by	shooting;	
forfeiture;	driver	license	
revocation	

	
625	

77.9%	

21.6%	

0.5%	

851.5%	

Use	of	wire	
communication	or	
electronic	communication	
in	drug	related	
transactions	

	
	
134	
78.4%	

20.1%	

1.5%	

787.6%	

Offense

Smuggling	
Tampering	w/	a	public	
record	
Marijuana	Violation	

Kidnapping	
Illegal	control	of	an	
enterprise;	illegally	
conducting	an	enterprise	

Number	of	
US	
incarcerations Citizens

239	
45	

Narcotic	Drug	Violation	

16,804	

79.6%	

20.1%	

0.4%	

784.8%	

Aggravated	Driving/DWI	

5,263	

80.9%	

19.1%	

0.0%	

739.5%	

20,615	

79.4%	

18.8%	

1.8%	

729.3%	

Aggravated	DUI	

36
Conspiracy	

274	

81.4%	

18.2%	

0.4%	

703.9%	

Discharging	a	firearm	at	a	
structure	

	
244	

82.0%	

18.0%	

0.0%	

694.4%	

Duty	to	give	information	
and	assistance;	alcohol	or	
other	drug	screening	

	
	
62	

82.3%	

17.7%	

0.0%	

681.6%	

1,950	

83.4%	

15.8%	

0.8%	

595.8%	

Facilitation	

45	

84.4%	

15.6%	

0.0%	

585.3%	

Cont.	sexual	abuse	of	a	
child	

26	

84.6%	

15.4%	

0.0%	

577.7%	

	
397	

84.4%	

14.6%	

1.0%	

543.6%	

16,774	

84.9%	

14.5%	

0.5%	

540.5%	

	
325	

84.6%	

14.5%	

0.9%	

537.1%	

21	

85.7%	

14.3%	

0.0%	

529.3%	

	
	
	
	
	
401	

85.5%	

14.2%	

0.2%	

526.2%	

1,115	

85.1%	

14.2%	

0.7%	

524.2%	

nd

1,159	

85.2%	

13.9%	

0.9%	

512.0%	

st

1,513	

85.6%	

13.9%	

0.5%	

511.4%	

Burglary	in	the	1 	degree	

1,391	

85.6%	

13.5%	

0.9%	

495.4%	

Molestation	of	a	child	

2,069	

86.7%	

13.0%	

0.3%	

472.8%	

Possess,	Sell,	Marijuana	

707	

87.4%	

12.4%	

0.1%	

448.3%	

Sexual	Abuse	

999	

87.7%	

12.1%	

0.2%	

433.6%	

87.2%	

12.1%	

0.8%	

431.1%	

535	

86.5%	

12.0%	

1.5%	

427.0%	

8,660	

87.6%	

11.9%	

0.4%	

425.5%	

144	

88.2%	

11.8%	

0.0%	

420.1%	

87.0%	

11.6%	

1.4%	

410.8%	

Manslaughter	

Discharge	Firearm	in	City	
Limit	
Dangerous	Drug	Violation	
Participating	in	or	assisting	
a	criminal	syndicate	
Unsworn	Falsification	
Accidents	involving	death	
or	physical	injuries;	failure	
to	stop;	driver	license	
revocation;	restricted	
privilege	to	drive;	alcohol	
or	other	drug		
Sexual	Assault	
2 	degree	murder	
1 	degree	murder	
st

Promoting	prison	
contraband;	exceptions;	xradiation;	body	scans	
Negligent	Homicide	
Armed	Robbery	
Stalking	
DUI	
Liquor/Drugs/VPRS/Combo	

2,397	

69	

37
Solicitation	

35	

88.6%	

11.4%	

0.0%	

403.5%	

Endangerment	

4,402	

87.9%	

11.3%	

0.7%	

398.4%	

Taking	identity	of	another	
person	or	entity	

1,010	

88.0%	

11.0%	

1.0%	

384.1%	

87.8%	

11.0%	

1.2%	

383.5%	

88.4%	

10.9%	

0.7%	

379.5%	

87.6%	

10.8%	

1.6%	

376.2%	

87.7%	

10.8%	

1.5%	

374.4%	

Dangerous	or	deadly	
assault	by	prisoner	or	
juvenile	

82	

Sexual	Exploitation	of	a	
Minor	

294	

Participating	in	or	assisting	
a	criminal	street	gang	

185	

Involving	or	using	minors	
in	drug	offenses	

65	

Disorderly	Conduct	

2,313	

89.1%	

10.3%	

0.6%	

355.2%	

Possession	&	Sale	Narcotic	

1,482	

89.7%	

10.3%	

0.0%	

351.8%	

49	

89.8%	

10.2%	

0.0%	

349.5%	

89.4%	

10.2%	

0.5%	

347.4%	

89.5%	

10.1%	

0.3%	

346.7%	

90.0%	

10.0%	

0.0%	

340.6%	

89.4%	

9.9%	

0.7%	

337.9%	

89.6%	

9.9%	

0.5%	

336.7%	

89.7%	

9.9%	

0.4%	

335.6%	

90.2%	

9.8%	

0.0%	

331.9%	

90.2%	

9.8%	

0.0%	

331.0%	

476	

90.3%	

9.7%	

0.0%	

325.7%	

Trafficking	Identity	

83	

89.2%	

9.6%	

1.2%	

324.6%	

Attempt	to	Commit	

21	

90.5%	

9.5%	

0.0%	

319.6%	

Flight	from	Law	in	A	
Vehicle	

3,055	

90.0%	

9.4%	

0.6%	

315.3%	

st

Escape	1 	Degree	
Sexual	Conduct	with	a	
Minor	

2,629	

Theft	Means	of	
Transportation	

7,407	

DWI	License	SuspendRevoke	

4,799	

Possession	of	burglary	
tools	

2,706	

Aggravated	Assault	
Unlawful	use	of	means	of	
transportation	

26,019	
6,634	

Interference	with	
monitoring	devices	

51	

Taking	identity	of	another	
person	or	entity	

92	

Unlawful	Imprisonment	

38
Misconduct	involving	
weapons	
Driving	While	Intoxicated	
Riot	
DWI	Liquor	or	Drugs	
Luring	a	minor	for	sexual	
exploitation	
nd

7,413	

90.2%	

9.2%	

0.6%	

305.9%	

864	

91.0%	

9.0%	

0.0%	

297.7%	

68	

89.7%	

8.8%	

1.5%	

288.7%	

274	

91.2%	

8.8%	

0.0%	

285.9%	

89.7%	

8.6%	

1.7%	

279.8%	

58	

Burglary	in	the	2 	degree	

10,249	

91.6%	

8.1%	

0.3%	

255.0%	

Theft	

25,933	

91.9%	

7.9%	

0.2%	

249.9%	

6,657	

92.1%	

7.7%	

0.2%	

240.1%	

92.2%	

7.7%	

0.2%	

237.7%	

92.2%	

7.5%	

0.3%	

228.9%	

91.0%	

7.5%	

1.5%	

228.8%	

Forgery	
Possession	and	sale	of	a	
vapor-releasing	substance	
containing	a	toxic	
substance	
rd

Burglary	in	the	3 	degree	
Aggravated	taking	identity	
of	another	person	or	entity	

	
	
600	
11,840	
268	

Securing	the	proceeds	of	
an	offense	

	
41	

90.2%	

7.3%	

2.4%	

222.3%	

Prisoner	assault	with	
bodily	fluids	

	
69	

92.8%	

7.2%	

0.0%	

219.2%	

	
2,049	

92.6%	

7.0%	

0.3%	

209.6%	

328	

92.7%	

7.0%	

0.3%	

208.9%	

86	

93.0%	

7.0%	

0.0%	

207.3%	

92.4%	

6.9%	

0.6%	

205.8%	

Criminal	possession	of	a	
forgery	device	
Criminal	Simulation	
Child	Prostitution		
Aggravated	criminal	
damage	

2,996	

Computer	Fraud	

58	

93.1%	

6.9%	

0.0%	

203.8%	

Threat-Intimidate	

394	

92.4%	

6.9%	

0.8%	

201.9%	

Arson	of	
Structure/Property	

321	

93.1%	

6.5%	

0.3%	

188.2%	

Hindering	prosecution	in	
the	first	degree	

245	

93.1%	

6.5%	

0.4%	

187.7%	

92.5%	

6.5%	

1.1%	

184.2%	

92.1%	

6.4%	

1.5%	

183.5%	

Tamper	with	Physical	
Evidence	
Failure	to	appear	in	the	

93	
202	

39
first	degree	
Cruelty	to	animals	

47	

Fraud	&	False	Statements	
to	Obtain	Benefit	

47	

Obstruct	Criminal	
investigation		

47	

Public	Sexual	Indecency	

	

174	

91.5%	

6.4%	

2.1%	

181.2%	

93.6%	

6.4%	

0.0%	

181.2%	

93.6%	

6.4%	

0.0%	

181.2%	

93.7%	

6.3%	

0.0%	

178.5%	

40
	
Comparison	to	Federal	data	
	
About	13%	of	prisoners	are	in	the	federal	prison	system,29	and	it	would	be	a	mistake	to	
use	state	prison	systems	to	infer	much	about	the	federal	system	as	the	types	of	crimes	
committed	by	federal	prisoners	differ	dramatically	from	those	in	state	prisons.		At	the	
end	of	2017,	there	were	a	total	of	37,557	confirmed	non-U.S.	citizens	under	the	
supervision	by	the	Federal	Bureau	of	Prisons	and	35,334	(or	92	percent)	were	
undocumented	immigrants.30			Undocumented	immigrants	make	up	almost	one-fifth	of	
Federal	inmates,	substantially	more	than	their	share	of	those	in	Arizona	state	prisons,	
though	differences	in	what	is	covered	by	federal	and	state	law	and	different	emphases	
on	enforcement	explain	the	gap.		Part	of	the	higher	rate	is	simply	because	the	fourth	
largest	category	of	people	are	in	federal	prison	for	are	immigration	violations	(7.3%	on	
November	25,	2017).31	
	
There	are	other	major	differences	in	the	composition	of	federal	and	state	cases.		For	
example,	46.3%	of	federal	prisoners	have	drug	offenses	as	their	most	serious	offense,	
while	that	is	true	for	only	22.2%	of	Arizona	state	prisoners.		The	difference	between	
federal	and	state	drug	offenders	extends	well	beyond	the	number	of	cases,	as	the	
federal	government	tends	to	concentrate	on	more	serious	offenders.			
	
Another	major	difference	involves	violent	crime:	41.4%	of	Arizona	prisoners	most	
serious	offense	involved	violent	crime,	but	that	is	true	for	only	16.1%	of	federal	
prisoners.		To	get	an	idea	of	how	large	this	difference	actually	is,	in	2014,	there	were	
317	murder	incarcerations	in	Arizona	and	only	124	in	the	entire	federal	system.		Again,	
the	crimes	for	federal	and	state	offenders	differ	in	other	ways,	such	as	where	the	crime	
occurred	(e.g.,	on	federal	property	or	Indian	reservations).	

29

	E.	Ann	Carson	and	Elizabeth	Anderson,	“Prisoners	in	2015,”	Office	of	Justice	Programs,	US	Department	
of	Justice,	December	2016	(https://www.bjs.gov/content/pub/pdf/p15.pdf).	
30
	Department	of	Homeland	Security,	“Alien	Incarceration	Report	Fiscal	Year	2017,	Quarter	4,”	
Department	of	Justice,	December	21,	2017	
(https://www.dhs.gov/sites/default/files/publications/Alien_Incarceration_Report_OIS_FY17_Q4_2.pdf).		
Federal	Bureau	of	Prisons	(https://www.bop.gov/about/statistics/population_statistics.jsp).	
31
	Federal	Bureau	of	Prisons	(https://www.bop.gov/about/statistics/statistics_inmate_offenses.jsp).	

41

	
Table	12:	Comparing	shares	of	Undocumented	Immigrants	in	Federal	and	State	
Prison	Systems	
	
US	Citizens	

Undocumented	
immigrants:	non-US	
citizen,	Not	Legal	
Permanent	Resident	

Documented	
immigrants:	non-US	
citizen,	Legal	
Permanent	Resident	

79.6%	

19.2%	

1.2%	

Arizona	Department	 88.1%	
of	Corrections	
January	to	June	
2017	

11.2%	

0.7%	

Arizona	Department	 84.3%	
of	Corrections	
January	1985	
through	June	2017	

12.96%	

1.75%	

Federal	Bureau	of	
Prison	

	
	
Unfortunately,	the	U.S.	Sentencing	commission	rarely	breaks	down	the	crimes	for	nonU.S.	citizens	by	whether	they	are	in	the	country	legally	or	not.		Table	13	shows	the	
breakdown	for	Arizona’s	crimes	by	both	the	total	non-citizens	and	by	whether	they	are	
in	the	country	legally.		The	table	shows	clearly	that	there	are	frequently	huge	
differences	between	the	federal	and	state	prison	system	in	terms	of	the	share	of	crimes	
committed	by	non-citizens,	with	kidnapping	and	drug	possession	being	the	two	largest.

42

	
	
Table	13:	Percent	of	crimes	committed	by	non-citizens	in	2014	
	

Federal*	

Arizona	

Crime	

Non-citizen	(Undocumented	
immigrants	and	documented	
immigrants)	

Arizona	
Undocumented	
immigrants:	nonUS	citizen,	Not	
Legal	Permanent	
Resident	

Documented	
immigrants:	
non-US	citizen,	
Legal	
Permanent	
Resident	

Murder/	
8.9%	
Manslaughter	

14.2%	

11.4%	

2.8%	

Kidnapping	

40%	

11.6%	

10.7%	

0.8%	

Sexual	Abuse	 7.5%	

17.3%	

13.3%	

4%	

Robbery	

2.9%	

9.7%	

9.56%	

0.14%	

Drug	
possession	

80.5%	

26.3%	

24.96%	

1.35%	

Burglary	

0.0%	

6.5%	

5.8%	

0.66%	

*	U.S.	Sentencing	Commission,	Final	Quarterly	Data	Report,	Fiscal	Year	2014,	Table	26.	
	

43

	
Implications	for	National	Violent	Crime	Rate	
	
What	if	undocumented	immigrants	in	the	entire	country	committed	crime	at	the	same	
rate	that	they	do	in	Arizona?		Table	14	shows	the	number	of	violent	crimes	for	
undocumented	immigrants	if	they	committed	crime	at	the	same	rate	as	the	general	U.S.	
population	and	the	number	if	they	committed	them	at	the	same	rate	as	undocumented	
immigrants	in	Arizona.		Over	the	years	2014	to	2016,	there	would	have	been	an	annual	
average	of	about	946	more	murders	and	manslaughters,	5,218	rapes,	8,753	robberies,	
23,969	aggravated	assaults,	28,284	burglaries,	larceny	89,335,	and	vehicle	theft	19,330.		
For	murders,	that	would	represent	about	6%	more	than	would	otherwise	be	the	case.	
	
Table	14:	Estimated	crimes	if	undocumented	immigrants	commit	crime	in	the	rest	of	
the	United	States	as	they	are	in	Arizona	
Year	

Estimated	
Undocumented	
Immigrants	
Population	in	
millions	

Number	of	crimes	
committed	by	
Undocumented	
Immigrants	if	they	
committed	them	at	
same	rate	as	overall	
population	

Difference	in	
number	of	
crimes		

	
	493		
	544		
	603		
	
	
	2,954		
	3,128		
	3,348		
	
	
	11,239		
	11,247		
	11,617		
	
	

Number	of	crimes	
committed	by	
Undocumented	
Immigrants	if	they	
committed	them	at	
same	rate	as	Arizona	
Undocumented	
Immigrants	
	
	1,346		
	1,486		
	1,647		
Average	
	
	7,857		
	8,321		
	8,905		
Average	
	
	19,893		
	19,908		
	20,562		
Average	
	

Murder	
2014	
2015	
2016	
	
Rape	
2014	
2015	
2016	
	
Robbery	
2014	
2015	
2016	
	
Aggravated	
Assaults	
2014	
2015	
2016	
	
Burglary	
2014	
2015	
2016	

	
11.1	
11.0	
11.3	
	
	
11.1	
11.0	
11.3	
	
	
11.1	
11.0	
11.3	
	
	
11.1	
11.0	
11.3	
	
	
11.1	
11.0	
11.3	

	25,447		
	26,191		
	28,082		
	
	
	59,629		
	54,420		
	52,984		

	48,399		
	49,815		
	53,411		
Average	
	
	89,920		
	82,065		
	79,900		

22,953		
23,624		
25,330		
23,969		
	
30,291		
27,645		
26,916		

	
853		
942			
1,044		
946		
	
4,903		
5,193		
5,557		
5,218			
	
8,654		
8,660		
8,945		
8,753			
	

44

	
Larceny-Theft	
2014	
2015	
2016	
	
Vehicle	Theft	
2014	
2015	
2016	
	

	
	
11.1	
11.0	
11.3	
	
	
11.1	
11.0	
11.3	
	

	
	
	202,192		
	196,195		
	197,181		
	
	
	23,905		
	24,443		
	26,770		
	

Average	
	
	293,178		
	284,483		
	285,912		
Average	
	
	42,360		
	43,313		
	47,436		
Average	

28,284		
	
90,986		
88,288		
88,731		
89,335		
	
18,455		
18,870		
20,666		
19,330		

	
Conclusion	
	
To	the	extent	that	undocumented	immigrants	are	reluctant	to	report	crimes,	the	results	
underestimate	undocumented	immigrants	share	of	crimes.		Still,	even	given	this	caution,	
undocumented	immigrants	are	convicted	of	crimes	at	much	higher	rates	than	U.S.	
citizens	for	they	face	longer	sentences	for	more	serious	crimes.			
	
Our	reliance	on	incarceration	data	means	that	there	is	a	greater	confidence	in	the	
accuracy	of	whether	these	individuals	have	committed	crime.		But	it	also	means	that	we	
are	underestimating	the	number	of	crimes	and	social	costs	of	criminal	activity	by	
undocumented	immigrants.			
	
The	crime	rate	by	undocumented	immigrants	also	depends	on	their	estimated	share	of	
the	Arizona	population.		If	there	are	more	undocumented	immigrants	than	claimed,	
their	estimated	crime	rate	goes	down.		Of	course,	the	opposite	might	also	be	true.		Yet,	
the	confidence	intervals	associated	with	the	estimates	are	very	tight.	
	
By	lumping	together	documented	and	undocumented	immigrants,	previous	research	
has	missed	out	on	the	huge	differences	between	these	two	groups.		Just	as	
undocumented	immigrants	are	more	likely	to	be	criminals,	documented	immigrants	
tend	to	be	very	law-abiding.		The	results	here	also	show	that	legal	Hispanics	are	more	
law-abiding	than	the	average	Arizonan.		The	perception	that	Hispanics	are	more	likely	to	
be	criminals	arises	from	not	distinguishing	between	legal	and	illegal	Hispanics.	
	
Undocumented	immigrants	especially	commit	crimes	at	a	young	age	and	to	be	involved	
in	gangs,	but	then	they	suddenly	stop	being	convicted.			A	startling	difference	is	while	a	
large	percentage	of	U.S.	citizens	who	spend	their	lives	cycling	through	prison,	that	isn’t	
the	case	for	undocumented	immigrant	criminals.		The	implication	for	that	is	not	only	are	
the	crime	rates	by	undocumented	immigrants	higher	than	for	U.S.	citizens,	but	an	even	
larger	percentage	of	undocumented	immigrants	spend	at	least	part	of	their	lives	in	
prison.		The	question	that	can’t	be	definitely	answered	here	is	what	happens	to	these	

45

undocumented	immigrant	criminals.		Do	they	quickly	reform	after	being	in	prison	or	do	
they	leave	Arizona?		The	most	plausible	reading	of	the	evidence	suggests	that	they	are	
leaving	Arizona,	presumably	to	return	to	their	home	country,	overwhelmingly	Mexico.		

46

Bibliography	
	
Bohn,	Sarah,	Magnus	Lofstrom,	and	Steven	Raphael	Goldman,	“Did	the	2007?	Legal	Arizona	
Workers	Act	Reduce	the	State’s	Unauthorized	Immigrant	Population?”	Public	Policy	
Institute	of	California,	November	2012.	
Butcher,	Kristin	F.,	and	Anne	Morrison	Piehl,	“Why	Are	Immigrants’	Incarceration	Rates	So	
Low?	Evidence	on	Selective	Immigration,	Deterrence,	and	Deportation,”	NBER	Working	
Paper	no.	13229,	National	Bureau	of	Economic	Research,	July	2007.	
Ewing,	Walter	A.,	Daniel	Martinez,	and	Ruben	Rumbaut,	“The	Criminalization	of	Immigration	
in	the	United	States,”	American	Immigration	Council	Special	Report,	July	2015.		
Hagan,	John,	and	Alberto	Palloni.	1998.	Immigration	and	crime	in	the	United	States.	In	The	
Immigration	Debate,	eds.	James	P.	Smith	and	Barry	Edmonston.	Washington,	DC:	
National	Academy	Press.		
Hickman,	Laura	J.,	and	Marika	J.	Suttorp.	2008.	Are	deportable	aliens	a	unique	threat	to	
public	safety?	Comparing	the	recidivism	of	deportable	and	nondeportable	aliens.	
Criminology	&	Public	Policy	7:59–82.		
Landgrave,	Michelangelo,	and	Alex	Nowrasteh,	“The	DREAMer	Incarceration	Rate,”	
Immigration	Research	and	Policy	Brief,	Cato	Institute,	August	30,	2017.	
	
Lott,	John	R.,	Jr.,	and	Rujun	Wang,	“Prisoners	in	Arizona:	A	2017	Update	on	Selected	Topics,”	
Arizona	Prosecuting	Attorneys’	Advisory	Council,	November	30,	2017.	
	
McCormick,	Robert,	and	Robert	Tollison.	1984.	“Crime	on	the	Court,”	Journal	of	Political	
Economy	92:	223-235.	
	
Roberts,	Bryan	W.,	“Illegal	Immigration	Outcomes	on	the	U.S.	Southern	Border,”	Cato	
Journal,	Fall	2017.	
	
Rumbaut,	Rubén	G.	,	and	Walter	Ewing,	The	Myth	of	Immigrant	Criminality	and	the	Paradox	
of	Assimilation:	Incarceration	Rates	among	Native	and	Foreign-Born	Men	(Washington:	
DC:	Immigration	Policy	Center,	American	Immigration	Law	Foundation,	Spring	2007.	
Salas-Wright,	Christopher	P.,	Michael	Vaughn,	and	Trenette	Clark	Goings.	2017.	“Immigrants	
from	Mexico	experience	serious	behavioral	and	psychiatric	problems	at	far	lower	rates	
than	US-born	Americans,”	Social	Psychiatry	and	Psychiatric	Epidemiology	52:	1325–1328.	
Stowell,	Jacob,	Steven	Messner,	Kelly	McGeever,	and	Lawrence	Raffalovich.	2009.	
“Immigration	and	the	recent	violent	crime	drop	in	the	United	States:	A	pooled,	cross-

47

sectional	time-series	analysis	of	metropolitan	areas,”	Criminology	47:889-928.	
	
Waters,	Mary	and	Maris	Gerstein	Pineau,	Editors,	2015.	The	Integration	of	Immigrants	into	
American	Society.	National	Academies	Press:	Washington	DC.		
	

48

Appendix	1:	Redoing	Table	1	by	individual	rather	than	looking	at	each	incarceration	
	

Appendix	1	breaks	down	the	data	per	person	rather	than	per	incarceration	as	in	Table	1.		These	results	
show	a	43%	higher	share	of	undocumented	immigrants.		There	were	5,322	individuals	who	were	not	
included	in	this	breakdown	because	their	citizenship	status	changed	between	incarcerations.		The	16%	
share	of	incarcerations	from	1985	to	2017	implies	that	undocumented	immigrants	were	convicted	at	least	
233%	more	often	than	Arizonans	in	general.			

	
Table	A1:	Race	and	Citizenship	of	those	incarcerated	in	Arizona	by	person	incarcerated
1985	to	June	2017

US	Citizens

Undocumented	immigrants:	
Non-US	citizen,	Not	Legal	
Permanent	Resident

Documented	immigrants:	NonUS	citizen,	Legal	Permanent	
Resident

Caucasian 108,980

43.0%

0.4%

0.1%

AfricanAmerican

26,647

10.3%

0.3%

0.1%

Native	
American

15,292

6.1%

0.0%

0.0%

Hispanic,	
US	Born† 56,500

22.2%

0.3%

0.0%

Hispanic,	
not	US	
born
38,668

0.7%

14.2%

0.6%

Asian

770

0.2%

0.1%

0.0%

Other

3,317

0.6%

0.6%

0.1%

Unknown	
race
131

0.0%

0.0%

0.0%

Total

83.2%

16.0%

0.9%

RACE

Total*

250,305

*	Totals	are	for	those	who	can	be	classified	as	U.S.	citizen,	undocumented	immigrants,	or	documented	immigrants.	
†	One	concern	with	the	table	is	that	there	are	US	born	individuals	who	are	listed	as	not	being	US	citizens	or	who	are	attempting	to	
gain	citizenship	(though	this	last	group	is	extremely	small).		We	asked	Bill	Montgomery,	the	County	prosecutor	for	Maricopa	county,	
and	the	Arizona	Department	of	Corrections	about	these	cases,	and	we	were	told	that	they	involved	people	renouncing	their	US	
citizenship	and	then	returning	to	the	US	or	whose	US	citizenship	was	never	claimed.		Given	the	small	number	of	people	who	have	
renounced	their	citizenship,	this	assumption	seems	questionable.	(https://www.federalregister.gov/documents/2017/11/02/201723885/quarterly-publication-of-individuals-who-have-chosen-to-expatriate-as-required-by-section-6039g).		Possibly	they	are	
recording	errors,	where	non-citizens	are	listed	as	born	in	the	US	or	people	born	in	the	US	are	listed	as	non-citizens.		In	any	case,	the	
number	of	these	cases	is	very	small	and	does	not	appreciably	alter	the	results	presented	here	(See	Appendix	A3).

49

	
	
Appendix	2:	Relative	shares	of	Arizona	population	and	prison	population	by	race	
	
By	racial	group,	Asians	had	by	far	the	smallest	share	of	their	population	who	were	
convicted	of	a	crime	in	2014.		Overall,	Hispanics	were	less	likely	to	be	convicted	of	a	
crime	in	2014	than	either	African-Americans	or	Native	Americans.		Given	the	high	rate	
that	undocumented	Hispanic	immigrants	are	convicted	of	crime,	the	share	of	crime	
committed	by	all	other	Hispanics	appears	to	be	very	low.		Undocumented	Hispanic	
immigrants	account	for	about	8.8	percentage	points	of	the	state	population	and	about	
61.5%	of	the	Hispanic	share	of	the	prison	population.		Thus	that	means	the	U.S.	citizen	
portion	of	Hispanics	(28.8%)	is	almost	twice	the	remaining	prison	population.	Note	that	
23.14%=8.8%(Non-US	citizen	Hispanics)*2.63(the	higher	crime	rate	for	undocumented	
immigrants).			
	

Table	A2:	Comparing	shares	of	the	Arizona	Population	to	shares	of	Arizona	Prison	Population	by	Race
RACE

Percent	of	the	Arizona	
Population

Percent	of	Arizona	
Prison	Population

Ratio	(Percent	Prison	
Population/	Percent	
Entering	Arizona	
Population)

Caucasian

55%

40.4%

0.735

African-American

4.62%

12.2%

2.641

Native	American

5.15%

6.95%

1.35

Hispanic

30.48%

38.61%

1.267

Asian

3.19%

0.32%

0.100

50

Appendix	3:	Examining	to	see	if	the	estimates	are	sensitive	to	cases	where	US	born	
people	were	listed	as	either	undocumented	or	documented	immigrants.	
	
As	noted	earlier,	while	it	is	possible	that	some	of	the	convicts	who	are	listed	as	US	born	
and	as	either	undocumented	or	documented	immigrants	renounced	their	citizenship	or	
never	had	their	citizenship	claimed	for	them,	it	is	also	possible	that	there	are	data	
errors.		Even	if	all	these	cases	are	in	error,	the	number	of	cases	is	still	small	and	they	
don’t	appreciably	alter	the	results.		To	see	this,	we	recalculated	the	initial	numbers	on	
page	6	and	those	shown	in	Table	3	by	removing	all	cases	where	convicts	were	classified	
as	undocumented	or	documented	immigrants	as	well	as	US	born.	
	
To	see	how	sensitive	the	results	are	to	the	inclusion	of	these	observations.	The	Arizona	
data	show	that	undocumented	immigrants	account	for	11.3%	of	convictions	for	first	and	
second	most	serious	offenses	(10.7%	of	the	most	serious	offenses	and	13.0%	of	the	
second	most	serious	offenses).		That	represents	just	a	0.5	percentage	point	change	from	
what	we	reported	earlier	in	undocumented	immigrants	share	of	criminal	convictions.	
	
Table	A3:	Share	of	Prisoners	for	Arizona	by	Most	Serious	Current	Offense	(Cases	where	there	are	at	20	
incarcerations	for	1985	to	June	2017):	Excluding	those	cases	where	the	documented	or	undocumented	
immigrant	had	also	been	listed	as	US	born	

Offense

Smuggling
Compounding	crime

Number	of	
US	
incarcerations	 Citizens

310	

20.6%

78.1%

1.3%

1,526.3%

22	

50.0%

50.0%

0.0%

941.7%

Unlawful	copying	or	
sale	of	sounds	or	
images	from	recording	
devices

27	

Tampering	w/	a	public	
record

63	

Marijuana	Violation

Percent	increase	in	the	
Undocumented	 Documented	
undocumented	
immigrants:	
immigrants:	
immigrants	convicted	
non-US	citizen,	
non-US	
relative	to	their	average	
Not	Legal	
citizen,	Legal	
share	of	the	Arizona	
Permanent	
Permanent	
State	Population	from	
Resident
Resident	
1985	to	2017

	
	

30,208	

	
	

	
	

	
	

51.9%

48.1%

0.0%

903.1%

58.7%

38.1%

3.2%

693.7%

66.4%

32.6%

1.0%

579.3%

69.1%

30.1%

0.8%

526.9%

Criminal	
Impersonation

977	

Money	Laundering

295	

64.1%

29.2%

6.8%

507.3%

3,160	

72.3%

26.9%

0.8%

459.7%

Kidnapping

51

Illegal	control	of	an	
enterprise

	

	

	

	

549	

73.6%

24.8%

1.6%

416.1%

Theft	by	extortion

92	

77.2%

22.8%

0.0%

375.5%

Drive	by	shooting;	
forfeiture;	driver	
license	revocation

668	

78.4%

21.1%

0.4%

339.7%

Discharging	a	firearm	
at	a	structure
Duty	to	give	
information	and	
assistance;	alcohol	or	
other	drug	screening

294	

83	

204	

Conducting	a	chop	
shop

91	

Solicitation
Conspiracy

18.1%

282.7%

	
	
0.0%

	
	

276.5%

	
	

	
	

1.0%

267.6%

82.4%

17.6%

0.0%

266.3%

47	

83.0%

17.0%

0.0%

254.6%

417	

83.5%

16.1%

0.5%

234.7%

84.1%

15.9%

0.0%

231.1%

81.8%

15.7%

2.5%

227.5%

83.1%

15.6%

1.3%

224.7%

85.0%

14.6%

0.4%

204.2%

41,243	
77	
31,949	
470	

84.3%

	

14.3%

	
	
	
	
	
	
562	

84.9%

	

1.5%

	
	
	
	
	
	
14.2%

	

197.0%

	
	
	
	
	
	
0.9%

	
	

	
	

17.6%

Aggravated	DUI

Accidents	involving	
death	or	physical	
injuries;	failure	to	
stop;	driver	license	
revocation;	restricted	
privilege	to	drive;	
alcohol	or	other	drug	
screening

0.0%

	
	

81.4%

10,452	

Discharge	Firearm	in	
City	Limit

81.9%

	
	

	
	

	
	

Aggravated	
Driving/DWI

Narcotic	Drug	
Violation

18.4%

	
	

Use	of	wire	
communication	or	
electronic	
communication	in	
drug	related	
transactions	

Continuous	sexual	
abuse	of	a	child

81.6%

	
	

	

	
	
	
	
	
	
196.6%

52

Manslaughter
Facilitation
Participating	in	or	
assisting	a	criminal	
syndicate
Sexual	Assault

2,834	

85.2%

13.9%

0.9%

189.6%

59	

86.4%

13.6%

0.0%

182.5%

	
428	
1,706	

	

	

	

84.6%

13.6%

1.9%

182.3%

85.9%

13.0%

1.1%

171.1%

86.1%

12.9%

0.9%

169.8%

st

Burglary	in	the	1 	
degree

1,753	

nd

2,204	

86.6%

12.8%

0.5%

167.5%

1 	degree	murder

st

1,790	

86.5%

12.6%

0.8%

163.0%

Narcotic	PossessTransport

24	

87.5%

12.5%

0.0%

160.4%

86.9%

12.1%

1.0%

152.5%

87.9%

12.0%

0.1%

150.5%

87.6%

11.7%

0.6%

144.3%

2 	degree	murder

Molestation	of	a	child

4,885	

Possess,	Sell,	
Marijuana

865	

Sexual	Abuse

2,021	

Keeping	or	residing	in	
house	of	prostitution;	
employment	in	
prostitution

	
	
	
27	

	
	
	

	
	
	

	
	
	

88.9%

11.1%

0.0%

131.5%

88.2%

11.1%

0.7%

130.8%

Dangerous	Drug	
Violation

32,665	

Armed	Robbery

10,493	

88.7%

10.9%

0.4%

127.3%

28	

89.3%

10.7%

0.0%

123.2%

757	

88.0%

10.7%

1.3%

122.9%

88.8%

10.6%

0.6%

121.2%

88.7%

10.6%

0.7%

119.9%

88.5%

10.4%

1.1%

117.6%

89.6%

10.4%

0.0%

117.4%

88.6%

10.2%

1.1%

113.1%

Unsworn	Falsification
Negligent	Homicide
Sexual	Conduct	with	a	
Minor

4,597	

Promoting	prison	
contraband;	
exceptions;	xradiation;	body	scans

3,250	

Endangerment

6,798	

Possession	&	Sale	
Narcotic

2,070	

Interference	with	
monitoring	devices

88	

53

DWI	License	SuspendRevoke

7,525	

DWI	Liquor	or	Drugs

500	

Involving	or	using	
minors	in	drug	
offenses

121	

Dangerous	or	deadly	
assault	by	prisoner	or	
juvenile

101	

Taking	identity	of	
another	person	or	
entity

1,841	

DUI	
Liquor/Drugs/VPRS/Co
mbo

173	

Possession	of	burglary	
tools

3,503	

89.9%

10.1%

0.0%

110.1%

90.0%

10.0%

0.0%

108.3%

89.3%

9.9%

0.8%

106.6%

89.1%

9.9%

1.0%

106.3%

89.2%

9.9%

0.9%

106.0%

88.4%

9.8%

1.7%

104.7%

89.8%

9.4%

0.8%

95.1%

89.3%

9.3%

1.3%

94.4%

Participate	Criminal	
Street	

225	

Child	Prostitution

129	

89.9%

9.3%

0.8%

93.8%

Escape	1st	Degree

65	

90.8%

9.2%

0.0%

92.3%

38,181	

90.3%

9.1%

0.6%

90.2%

90.6%

9.0%

0.4%

87.4%

91.0%

9.0%

0.0%

86.6%

91.1%

8.9%

0.0%

85.0%

90.7%

8.9%

0.5%

84.5%

91.2%

8.8%

0.0%

82.5%

90.9%

8.6%

0.5%

79.5%

91.1%

8.5%

0.4%

77.3%

89.5%

8.4%

2.1%

75.4%

Aggravated	Assault		
Unlawful	use	of	means	
of	transportation
Obstruction	of	A	
Criminal	Investigation
DWI

8,461	
67	
1,408	

Sexual	Exploitation	of	
a	Minor

847	

Unlawful	
Imprisonment

685	

Flight	from	Law	in	a	
Vehicle	
Theft	Means	of	
Transportation
Assault	

4,202	
10,425	
95	

54

Criminal	Simulation	

91.4%

8.4%

0.2%

75.1%

91.1%

8.4%

0.5%

74.1%

3,559	

91.0%

8.2%

0.8%

71.5%

Securing	the	proceeds	
of	an	offense

49	

87.8%

8.2%

4.1%

70.1%

Prescription-only	drug	

141	

92.2%

7.8%

0.0%

62.5%

13,617	

91.9%

7.8%

0.3%

61.6%

323	

92.3%

7.7%

0.0%

61.2%

Misconduct	involving	
weapon	
Disorderly	Conduct	

nd

Burglary	2 	Degree
Stalking
Aggravated	taking	
identity	of	another	
person	or	entity

	

452	
10,545	

	

	

	
488	

91.4%

	

	
7.6%

	

	
1.0%

	
58.0%

55

Appendix	4:	Accounting	for	other	Factors	by	Race	and	Citizenship	
Table	A4:	Accounting	for	other	factors	when	looking	at	differences	by	race	and	citizenship,	using	the	Arizona	Department	of	
Corrections	data	from	1985	to	2017	(Logit	with	fixed	effects	for	year	of	admission	to	prison	and	jurisdiction	they	committed	
the	crime,	absolute	z-statistics	in	parentheses)
Race / Race and Undocumented Immigrant
Control	
variable	

Caucasian	

Native	
Hispanic,	
American	 Not	U.S.	
citizen	

Hispanic,	 Hispanic	
Asian	
U.S.	
Undocumented	
citizen	
Immigrant	

Asian	
Undoc	
Immigrant	

-1.372	
(19.58)	
.000019	
(4.68)	

Caucasian	
AfricanAfricanUndocumented	 American	 American	
Immigrant	
Undocumented	
Immigrant	
Ommitted	
-3.17	
Ommitted	
(11.37)	
.00008	
-2.2e-06	 .000096	
(2.02)	
(0.39)	
(1.87)	

Gang	
Member	
Number	
of	Days	in	
Prison	
Age	
Admitted	
to	Prison	
Weapon	
Gun	
Weapon	
Knife	
Injured	
someone	
Male	

.9994	
(12.77)	
-.000282	
(25.33)	

0.341	
(5.17)	
0.0000526	
(8.10)	

1.223	
(24.80)	
9.02e-06	
(1.92)	

.2796	
(4.15)	
.000056	
(8.58)	

-.799	
(1.13)	
.00015	
(4.12)	

Ommitted	

.026	
(77.94)	

.0416	
(10.49)	

.0123	
(26.30)	

.0038	
(0.82)	

-.0026	
(3.87)	

-.0229	
(40.35)	

-.030	
(76.74)	

-.0271	
(45.92)	

-.0247	
(7.12)	

-.0255	
(3.46)	

-.415	
(24.48)	
-.276	
(10.87)	
-.136	
(9.78)	
-.4097	
(41.70)	
33648	

-.462	
(1.72)	
.201	
(0.71)	
.215	
(1.33)	
-.675	
(6.06)	
495.04	

.462	
(23.25)	
.006	
(0.17)	
.0556	
(2.96)	
.0449	
(3.10)	
17917	

-.052	
(0.24)	
-.8635	
(1.70)	
-.108	
(0.54)	
.585	
(3.27)	
186.30	

-.596	
(14.00)	
.504	
(12.58)	
.5139	
(21.00)	
-.421	
(23.96)	
17917	

-.0599	
(2.50)	
-.342	
(7.69)	
-.185	
(8.17)	
1.806	
(56.73)	
21487	

.248	
(15.69)	
.2297	
(8.79)	
.020	
(1.36)	
.141	
(12.31)	
20465	

-.053	
(2.18)	
-.3229	
(;7.18)	
-.1885	
(8.20)	
1.854	
(55.20)	
21448	

-0.79	
(0.55)	
-.163	
(0.69)	
0.225	
(1.86)	
-.496	
(6.10)	
361.97	

.1835	
(0.77)	
-.2076	
(0.45)	
.262	
(1.22)	
.843	
(2.83)	
198.7	

-3901	

-93168	

-3598	

-93168	

-128467	

-229500	

-124226	

-7006	

-1955.6	

ChiSquared	
Log	
-270719	
Likelihood	

.00014	
(2.47)