REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Adelman, 1970-2010 200 Metropolitan Time se- Violent and consistently associated R., Reid, L. Statistical Areas ries analysis property crime with fewer violent crimes and prop- W., Markle, (MSAs) exploring the rela- rates (UCR) erty crimes across the time period. G., Weiss, S., tionship between & Jaret, C. the percentage Each 1% increase in percentage for- (2017). of foreign-born eign-born was associated with 4.9 people (using cen- fewer violent crimes; 4.3 fewer rob- sus data) and beries; 0.11 fewer homicides; and violent/property 98.96 fewer property crimes. crime rates (us- ing Uniform Crime Report (UCR) data.

Allen, J., & 1990-2007 29 urban and ru- Analysis of the re- Juvenile prop- Negative but insignificant rela- Cancino, J. ral Texas counties lationship between erty crime tionship between the percentage M. (2012). situated along the the foreign-born (UCR) foreign-born population and juvenile Texas-Mexico border population and ju- property crime in both urban and ru- venile property ral counties. crime.

Amuedo- 2007-2014 All United States Ordinary least Changes in ar- When controlling for other variables, Dorantes, C., counties squares (OLS) re- rest and crime increases in resettlement Bansak, C., gression and rates (UCR) rates had no statistically significant & Pozo, S. instrumental vari- relationship with increases in crime (2018). able (IV) analysis rates when using either arrest or of- to explore the ef- fense data. fect of refugee resettlement on More complex analyses showed that crimes and arrests there is a statistically significant de- (UCR) crease in arrests for all crimes and violent crimes as refugee resettle- ment rates increase.

Immigration and Crime: A Public Policy Red Herring 1 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Bersani, B.E. 1997-2005 Individuals—8,984 Group-based tra- Self-reported First-generation immigrants were (2014) youth (includ- jectory modelling delinquent significantly less likely to have been ing 532 first- and (identifying clus- and criminal involved in any crime 988 second-gener- ters of individuals involvement ation immigrants) who display sim- in the previ- If first-generation immigrants were surveyed over ilar behavioral ous 12 months, involved in criminal offending, they nine annual waves trajectories over including prop- were also much more likely to have in the National a period of time) erty damage, desisted at an earlier age. Longitudinal Survey study examining theft, auto- Criminal involvement increased 1997 first-/second-gener- mobile theft, among second-generation immi- ation immigrants’ fraud, assault, grants, but there was no evidence involvement in and and selling that their rate of involvement or desistance from of- drugs patterns of offending were differ- fending over time ent from that of the native-born differed from those population. of native-born individuals Immigrant nationality groups did not act as a risk factor for offending.

Bersani, B. E. 1997-2005 Individuals — 4,122 Study exploring Self-reported Second-generation immigrants (2014). adolescents (3,533 whether predictors delinquent appeared to fall between their na- native-born and of crime amongst and criminal tive-born White and Black/Hispanic 589 second-gener- second-genera- involvement counterparts on the predictors of ation immigrants) tion immigrants in the previ- crime, including human and social surveyed over differ from those ous 12 months, capital, and environmental risk. nine waves in of native-born in- including prop- the National dividuals using erty damage, Second-generation immigrants Longitudinal Survey generalized hi- theft, auto- showed high degrees of conver- 1997 erarchical linear mobile theft, gence in all offending behavior when modeling fraud, assault, compared to White native-born and selling counterparts, including equivalent drugs risk and few significant differences in predictors of crime.

Second-generation immigrants showed similarities in less serious offending when compared to their native-born Black and Hispanic counterparts.

However, there was increased di- vergence in more serious offending behavior, including violent behavior and arrest, between second-gen- eration immigrants and their native-born Black and Hispanic counterparts.

CUNY Institute for State & Local Governance 2 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Bersani BE, 2000-2010 Individuals—1,354 Study comparing Arrest reports First-generation immigrants are less Loughran serious adolescent the relationship and self-report- likely to be involved in serious of- TA, Piquero offenders (1,061 na- between first-/sec- ed offending fending behavior and do not persist AR. (2014). tive-born, and 293 ond-generation in offending behavior as long as na- first- or second-gen- and the frequency tive-born peers. eration immigrants) of self-reported of- in the longitudi- fending behavior Second-generation immigrants on nal Pathways to and arrest average had higher risk of persistent Desistance Study offending over the life course when from Maricopa more assimilated and living in more County, AZ and disadvantaged contexts. Philadelphia County, PA

Bersani, B. E., 2000-2010 Individuals—1,354 Study using Self-reports High degree of correspondence & Piquero, A. serious adolescent descriptive, per- and official between self-reported arrest and of- R. (2017). offenders (and 293 mutation, and reports of ar- fending behavior and official reports first- or second-gen- hierarchical lin- rests across 22 of arrest within groups based on im- eration immigrants) ear model analyses offenses migrant generation. in the longitudi- exploring wheth- nal Pathways to er foreign-born First generation immigrant youth Desistance Study immigrants’ lower displayed lower levels of offending, from Maricopa levels of criminal both in self-reported data and official County, AZ and involvement is arrest records. Philadelphia the result of dif- There was no evidence of systemat- County, PA ferential crime ic crime reporting biases, or over or reporting practices under-reporting behavior, across im- between immi- migrant generations. grants and U.S. natives

Bucher, J., 2004 Individuals— in- Descriptive study Self-reported Over half of migrant workers report- Manasse, M., terviews of a exploring rates crime ed being the victims of theft (57%), & Tarasawa, convenience-based of victimization victimization about one in 6 were the victim of B. (2010). sample of 90 among undoc- robbery (16%), and one in ten (10%) undocumented im- umented male were the victim of a violent attack. migrant workers in migrant workers. Memphis, TN Crimes were only reported in one-quarter of cases (24%) and were only reported by the victim in one instance (the rest were by someone else, such as a supervisor/landlord).

Victimization was most common among recent immigrants.

Immigration and Crime: A Public Policy Red Herring 3 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Bui, H. N. 1994-1996 Individuals—12,868 Longitudinal study Self-reporting Higher immigrant generations (e.g., (2008). adolescents (in- on the associa- delinquent second- or third-generation) are as- cluding 962 tion between first/ behavior sociated with higher delinquency first-generation and second generation rates for some offenses among racial, 1,952 second-gen- status and self-re- ethnic and gender subgroups. eration) from the ported delinquent first two waves behavior of immi- of the National grant youth. Longitudinal Study of Adolescent to Adult Health

Chalfin, A. 1986-2004 46 MSAs Instrumental Index, violent, No significant relationship between (2014). variable analysis and property immigration and violent or property exploring the ef- crimes (UCR) crime rates in the included U.S. cit- fect of Mexican ies, with the exception of an increase immigration to the in the rate of assault. U.S. on crime rates (UCR data), using rainfall shocks as an instrumental variable on the ba- sis that very high or very low lev- els of rainfall in Mexico are likely to drive immigra- tion to the U.S.

Chalfin, A. 1980-2000 92 MSAs with a suf- Network instru- Indexed Immigration suggests a protective ef- (2015). ficient presence ment analysis crimes from fect on several crime types—a 1% of Mexican immi- exploring the ef- National Crime increase in the immigrant share grants for reliable fect of Mexican Victimization leads to a 13% reduction in rape, an estimation immigration rates Survey (NCVS) 11% reduction in larceny, and a 15% (U.S. Census data) data reduction in motor vehicle theft. on crimes report- ed to the police (NCVS data), using Mexican fertility rates as an instru- ment (Mexican immigrant share from the National Institute of Statistics and Geography)

CUNY Institute for State & Local Governance 4 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Chalfin, A., 2001-2009 Arizona after pas- Synthetic con- Index crimes, Violent crime fell 10% following the & Deza, M. sage of LAWA and trol analysis of violent crimes, implementation of LAWA and prop- (2020). 45 comparison the effect of Legal and property erty crime fell 20%. states that did not Arizona Workers crimes (UCR) pass comprehensive Act (LAWA) on There were large declines in the pop- E-Verify laws appli- crime rates. The ulation foreign-born Mexican males, cable to the private Legal Arizona between 15-24 years of age, with high sector Workers Act re- labor participation rates, after pas- quired Arizona sage of LAWA. employers to use Reduction in crime may be due to E-Verify to ensure the reduction in this young male all workers are au- population since young men have thorized to work in disproportionate involvement in the country. crime.

Ciancio, A. 2008-2014 3,181 U.S. counties Study analyzing Violent Democratic counties followed guide- (2017). enrolled in Secure the effects of fed- crimes, prop- lines by reducing immigration Communities pro- eral immigration erty crimes, enforcement while Republican coun- gram between and deportation homicides, and ties maintained higher levels of October 2008 and guidelines on local crime clear- enforcement. January 2013 crime and policing ance rates efficiency as mea- Democratic counties with higher sured by clearance non-citizen population shares saw rates. no increase in crime, and greater in- creases in clearance rates.

Overall, reducing immigration en- forcement did not result in increased crime, but did increase police effi- ciency (suggesting that the reduction of enforcement allowed police to fo- cus on solving more serious crimes, or resulted in non-citizens cooperat- ing with police).

Immigration and Crime: A Public Policy Red Herring 5 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Davies, G., 1990- All neighborhoods Analysis of the re- Homicide and The number of total immigrants in a & Fagan, J. 2000; in New York City lationship among assault rates neighborhood was a protective factor (2012). 2004-2008 immigration, lo- between 1990 for reductions in total, violent, prop- cal crime rates, and and 2000; ar- erty, and drug crime. patterns of police rests and street enforcement. stops from However, higher concentrations of 2004 to 2008 recent immigrants showed a slight increase in violent crime specifically.

Black immigrants tended to settle in places where social and econom- ic disadvantages were greatest, but cultural advantages yielded several protective factors that prevented in- creases in crime.

The effect of immigration on crime did not rest solely on race/ethnicity, but also cultural factors that impact- ed protective factors - both White and Black immigrant groups had strong insulating effects, but Latinx and Asian groups had no net impact overall, suggesting a need to fur- ther differentiate between different cultural groups within each race/ ethnicity.

Enforcement for total, violent, and property crimes was higher in plac- es with greater proportions of immigrants.

DiPietro, S. 1997-2002 Individuals—1,662 OLS analysis on Self-reported Overall, findings supported gender dif- M., & Cwick, children and ado- the association be- violent ferences in violent behavior based J. (2014). lescents, ages 9, 12 tween generation behaviors on level of assimilation and family and 15 years old (and status (first, first functioning. 714 first- through and a half, second second-generation) and third or more For boys, less assimilated generations and their prima- generations) and (first through second-generation im- ry caregivers, in self-reported vio- migrants) were less likely to report the longitudinal lent behaviors. violence than other native-born peers Project on Human in all models. This effect was more Development than twice as strong for first gener- in Chicago ation immigrants than for second Neighborhoods generation immigrants. For girls, less assimilated generations were less likely to report violence in the first model, but after controlling for family functioning, this relation- ship was reduced to non-significance.

CUNY Institute for State & Local Governance 6 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Graif, C., & 1995-2006; Chicago, IL Modeling Self-Reported Immigrant concentration either did Sampson, R. 2002-2006 spatial dynam- Violence not impact or decrease self-reported J. (2009). ics—geographically (Data from violence. Further, language diversity weighted regres- the Project was consistently negatively associat- sion study on on Human ed with self-reported violence. the relationship Development between the pro- in Chicago) portion of first generation immi- grants in Chicago and self-reported violence.

Green, D. 2012-2014 All U.S. states OLS regression UCR-reported Foreign population size had no as- (2016). and overdispersed rates including sociation with violent crime rates, Poisson regres- violent crime murder or rape except for the un- sion analysis on rates (as well documented Mexican population the relationship as subgroups showing a weak association at the between immi- of homicide 90% CI with overall violent crime. gration (American and rape); drug Community Survey crime arrest There was a small but significant as- (ACS) and violent rates (as well sociation with Mexican nationals, crime and drug ar- as subgroups and undocumented immigrants in rests (UCR) sales and particular, and drug arrest rates. possession)

Gunadi, C. 2006-2015 Individuals count- Estimation of the Incarceration Incarceration rate of undocu- (2019). ed in the American incarceration rate rate of imputed mented immigrants (~2%) was Community Survey of undocument- undocument- lower compared to U.S. natives (ACS) for estima- ed immigrants and ed immigrants (~3%), suggesting that undocument- tion of incarceration analysis of wheth- aged between ed immigrants were 33% less likely rates of imput- er undocumented 18 - 40 years, to be incarcerated compared to U.S. ed undocumented immigration con- vs. U.S. natives natives. citizens; all U.S. tributes to crime. (ACS); crimes states for immigra- per 100,000 No evidence that more recent im- tion-crime analysis (UCR) migrants were more likely to be institutionalized.

No evidence that undocumented im- migration contributed to violent or property crime.

Immigration and Crime: A Public Policy Red Herring 7 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Harris, C. T., 1999-2001 326 census places Analysis of the Violent in- Overall, recent Latinx immigra- & Feldmeyer, in California, New relationship be- dex arrest tion was unassociated with violence B. (2013). York, and Texas tween Latinx rates using a across all communities. immigration and 3-year aver- crime. Further, age of arrests In traditional destinations for im- the authors an- for the period migrants, recent Latinx immigration alyzed if the 1999-2001 yields a decrease in Latinx violence, association var- but unassociated with White and ies in traditional Black violence. vs. non-traditional In non-traditional destinations for immigrant desti- immigrants, recent Latinx immigra- nations or across tion was unassociated with White race/ethnic groups. violence, but did have a weak pos- itive association with Black and Latinx violence.

There were other key variables, in- cluding structural disadvantage, that appeared to have statistically sig- nificant relationships to increased violence across all models, with the exception of Latinx violence in non-traditional destinations.

Kubrin C.E., 1995-1996 Individuals—9,500 Longitudinal Number of dif- Community social capital and per- & Desmond (including 2,470 analysis of 9,500 ferent violent sonal victimization did not mediate S,A. (2015). first- or second- gen- adolescents testing acts committed the effects of immigrant concen- eration) adolescents potential mod- by adolescents. tration on adolescent violence in from the National erators, such as neighborhoods. Longitudinal Study community social of Adolescent Health capital or victim- ization, in the relationship be- tween immigration and crime.

CUNY Institute for State & Local Governance 8 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Kubrin, C. E., 2009-2011 2,740 census tracts Regression analy- Violent and Immigrant concentration was not Hipp, J. R., in the Southern sis exploring if the property crime significantly associated with neigh- & Kim, Y. A. California metropol- immigration-crime (UCR) borhood violent crime but was (2018). itan region nexus varies significantly associated with de- based on nation- creases in neighborhood property al origin, using four crime. models of increas- ing complexity: (1) When controlling for other factors, percentage foreign- neighborhoods with a higher per- born, (2) centage of Latinx immigrants had disaggregat- higher violent crime rates; neighbor- ing foreign-born hoods with a higher percentage of according to U.S. Black or Asian immigrants had lower racial categories, violent crime rates. (3) the world region The models with the most from which immi- explanatory power disaggregated im- grants originated, migrants by the region of the world and (4) categories of the sending country. organized based on immigrant Higher levels of immigrant con- groups’ tendency centration among immigrants to spatially co-lo- from West and North , North cate with others. America, and South , led to de- Data on foreign- creases in violent crime. born persons were obtained from the Higher levels of immigrant concen- U.S. Census, and tration among immigrants from violent and proper- Western, Eastern, and Central ty crime data were yielded increases in violent obtained from the crime. UCR. Higher levels of immigrant concen- tration among immigrants from West Africa and East and Southeast Asia yielded decreases in property crime.

Higher levels of immigrant concen- tration among immigrants from and the Caribbean led to increases in property crime.

Immigration and Crime: A Public Policy Red Herring 9 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Kubrin, C. E., 1999-2001 1,644 census tracts Series of regression Three-year av- In Chicago, neighborhoods with high & Ishizawa, in Chicago, IL and analyses on the re- erage count immigrant concentration had signif- H. (2012). Los Angeles, CA lationship between of homicides icant decreases in crime, even after neighborhood and robberies controls were included. immigrant concen- (UCR) tration and crime Immigrant neighborhoods, em- rates. bedded within larger immigrant communities, also had lower than average levels of violent crime.

In Los Angeles, neighborhoods with high immigrant concentration had a significant increase in crime un- til the addition of controls reversed the relationship, yielding significant decreases in crime. However, im- migrant neighborhoods, embedded within larger immigrant communi- ties, had higher than average levels of violent crime.

Landgrave, 2016 Individuals— im- Descriptive study Incarceration Incarceration rates for both doc- M., & putation of on the characteris- rate umented and undocumented Nowrasteh, undocumented im- tics of immigrants immigrants were much lower than A. (2018). migrant population who are incarcerat- native-born Americans. amongst inmates in ed. Study compared ACS data incarceration rates Specifically, undocumented immi- of native born grants were 49% less likely to be and foreign-born incarcerated, and documented im- people. migrants were 75% less likely to be incarcerated compared to incarcerat- ed native-born Americans.

Both documented and undocu- mented immigrants have a lower likelihood of incarceration within each race/ethnicity category com- pared to native-born Americans of the same race/ethnicity.

CUNY Institute for State & Local Governance 10 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Light, M. T., 1990-2014 All 50 states and Regression analysis Violent crime Higher rates of undocumented im- & Miller, T. Washington, DC on the associations (UCR and migration are generally associated (2018) between chang- NCVS) with lower rates of overall violent es in immigration crime in both law enforcement and and crime rates self-reported victimization data, but across all 50 states this relationship is not significant in and Washington, all cases. DC from 1990 to 2014. Data from Cross-sectional sub-analyses show the Center for higher rates of immigration are Migration Studies weakly associated with increases in and Pew Hispanic violent crime (i.e., longitudinal stud- Center were used ies should be given more weight). to determine the percentage of the population that was undocumented.

Lopez, K.M., 1994-2001 Individuals—763 Study investigating Instances of Mexican adolescents were less like- & Ventura Hispanic children whether offending self-report- ly to report violent offending when Miller, H. and adolescents among Hispanic ed offending compared to their other Hispanic (2011). (94% of whom adolescents varied of violent and adolescents. were first- or sec- by ethnicity and proper- ond-generation) in whether Hispanic ty crimes the longitudinal immigrants who Puerto Ricans adolescents were Project on Human were more accul- more likely to report violent offend- Development turated were more ing than other Hispanic adolescents. in Chicago likely to engage in Neighborhoods offending. When controlling for other crimi- nological variables, first-generation immigrants were significantly less likely to engage in delinquent behav- ior overall, especially relative to their third-generation counterparts.

Immigration and Crime: A Public Policy Red Herring 11 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Lyons, C. J., 1999 & 87 cities with popu- Multilevel instru- Homicide and Cities with a higher foreign-born Vélez, M. B., 2001 lations over 100,000 mental variable robbery rates proportion of the population have & Santoro, analyses looking lower crime rates. W. A. (2013). at the relationship between the for- The negative relationship between eign-born share of immigrant concentration and violent the population and crime rates was strengthened when homicide and rob- cities had more political opportu- bery rates. Authors nities for immigrants, though these looked at the mod- opportunities predicted rates of ho- erating effects of micide and robbery differently. immigrant politi- cal opportunities, representation in the police force, sta- tus, and politics.

MacDonald, 2000-2005 Census tracts in Los Analysis looking Total crimes In a number of different esti- J. M., Hipp, J. Angeles, CA at the relationship and violent mated models, greater predicted R., & Gill, C. between predicted crimes (UCR) concentrations of immigrants in (2013). immigrant con- neighborhoods are linked to signifi- centration (U.S. cant reductions in crime. Census) and neigh- borhood crime rates (UCR).

Martinez Jr., 1980-1989 Census tracts in Analysis of the Number of Overall, even across time and place, R., & Stowell, Miami, FL and San relationship be- homicide vic- higher numbers of immigrants J. I. (2012). Antonio, TX tween immigrant tims (internal did not lead to higher numbers of concentration on files stored in homicides. the number of Miami and homicides. San Antonio In Miami, neighborhoods with high- homicide inves- er levels of immigration had fewer tigation units), total homicides and drug-related ho- disaggregat- micides. For incidents disaggregated ed by subtypes: by ethnicity of the victim, there were escalation lower levels of Latinx homicides and homicides, non-Latinx White homicides. gang-motivat- Increases in the immigrant concen- ed homicides, tration in San Antonio showed no and drug relat- corresponding increases in total ho- ed killings micides or gang, drug, or escalation killings. In fact, heavily immigrant communities buffered gang warfare - areas with more immigrants had fewer gang-related homicides.

CUNY Institute for State & Local Governance 12 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Martinez Jr, 1980-2000 Neighborhoods in Longitudinal study Homicide Rates Overall, increases in the size of the R., Stowell, J. San Diego, CA looking at the as- and Social foreign-born population reduced le- I., & Lee, M. sociation between Disorganization thal violence over time. T. (2010) the concentration of foreign-born res- San Diego neighborhoods with high- idents, homicide er immigrant concentrations saw a rates, and social robust decline in homicides over- disorganization. all (and for White Non-Latinx and Latinx people in particular).

Change in concentration of immi- grants did not predict changes in homicides of Black people.

Neither immigration nor residen- tial instability (i.e., people moving in/out of residences) substantially contributed to increased neighbor- hood disorganization - instead social disorganization in immigrant cit- ies was likely related to economic deprivation.

Martínez- 1990-2010 107 U.S. cities Regression analysis Homicide and Adoption of a sanctuary policy had Schuldt, on the relationship robbery (UCR) no significant impact on homicides R. D., & between city-level but was associated with an 11% de- Martínez, D. sanctuary poli- cline in robberies. E. (2019). cies, concentration of unauthorized Increases in the relative proportion Mexican im- of the unauthorized Mexican immi- migrants, and grant population led to a decrease in homicide and rob- homicides (about 5% decrease for ev- bery incidents. ery 1 percentage point increase in representation), but only in cities with sanctuary policies.

Immigration and Crime: A Public Policy Red Herring 13 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Maynard, B. 2009-2010 Individuals— 12,642 Correlational study Bullying and Immigrant youth were significantly R., Vaughn, adolescents (1,068 looking at the rela- negative health more likely to experience at least one M. G., Salas- of which were born tionship between outcomes form of bullying, including bullying Wright, C. P., outside the United immigration sta- related to religious or racial factors, & Vaughn, S. States) in the nation- tus, bullying, and compared to native-born youth. R. (2016). ally representative negative health Health Behavior outcomes. Links between immigration and in School-Aged bullying victimization were also in- Children study fluenced by gender; first-generation immigrant boys were more likely to report bullying, including sexually related, religious, and cyber bullying, when compared to their native-born counterparts.

Bullied immigrant youth reported significantly poorer health and had a higher risk of being overweight com- pared to immigrant youth who did not report bullying.

Miles, T. J., 2004-2012 2,985 U.S. counties Differences- Violent and No meaningful reduction in the & Cox, A. B. in-differences property crime overall index crime rate or individual (2014). analysis of the ef- rates (UCR) violent offenses in counties partic- fect of Secure ipating in the Communities on program. crime rates. Secure Communities Overall, Secure Communities activa- was a deportation tion showed decreases in crime, but program which when county-specific trends were in- enabled automat- cluded, results suggested that Secure ic transmission of Communities activation led to more fingerprints taken crime. upon arrest to DHS The baseline decline was almost for verification of entirely due to reductions in the arrestees’ immigra- counties with the highest concentra- tion status. tions of foreign-born people.

Norris, C. 2000-2017 Arizona and 47 com- Synthetic control Violent and Following the SB 1070 implementa- (2020). parison states analysis of the ef- property crime tion, violent crime increased by 20%, fect of SB1070 on rates (UCR) aggravated assault increased by 10%, crime. and rape increased by 20%.

SB1070 was an Property crime did not change fol- Arizona law which lowing SB1070. extended immigra- tion enforcement power to local law enforcement.

CUNY Institute for State & Local Governance 14 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Nowrasteh, 2011-2017 Individual-level da- Descriptive study Homicide, sex Arrest and conviction rates were A. (2018) ta on arrests and comparing crime crime, larce- lower for immigrants compared to convictions of docu- rates among doc- ny, and other native- born people. mented immigrants, umented and crime rates undocumented undocumented Documented immigrants had lower immigrants, and na- immigrants, and arrest rates than undocumented im- tive-born people native-born people. migrants (though both groups had a in Texas, from the lower arrest rate than native-born). Texas Department of Homicide and sex crime conviction Public Safety rates were lower for both document- ed and undocumented immigrants compared to native-born people, though the difference was much greater for documented immigrants.

Larceny conviction rates were lower for all immigrants, but the differ- ence was greater for undocumented immigrants.

O’Brien, B. G., 2000-2015 4,055 U.S. cities Analysis of the re- Violent and No significant change in crime rates, Collingwood, lationship between property crime including violent, rape, and proper- L., & El- sanctuary city (UCR) ty crime, following sanctuary city Khatib, designation and vi- designation. S. O. (2019). olent and property crime. No significant difference in crime be- tween sanctuary cities and similarly situated cities without a sanctuary designation (matched based on popu- lation, size of the Latinx community, economic criteria, and other racial criteria).

Immigration and Crime: A Public Policy Red Herring 15 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Ousey, G. C., 1980-2010 156 large U.S. cit- Time series anal- Argument, felo- Overall, increases in immigration led & Kubrin, C. ies with populations ysis looking at ny, drug-related, to significant reductions in overall E. (2014). over 100,000 the relationship and gang-relat- homicide and drug-related homicide between immi- ed homicide. rates. gration and the (Supplementary prevalence of ho- Homicide A 10% increase in the immigration micide subtypes Reports (SHR)) measure was associated with a 3% disaggregated by decrease in the homicide rate. circumstance. There was no significant relationship between increases in immigration and other homicide subtypes.

Effects of changes in immigration vary by place with more pronounced decreases in overall and drug-relat- ed homicide rates in places with a higher initial (i.e. 1970) immigrant population base.

Portes, A., & 1992-2003 San Diego, CA Exploratory lon- Educational Likelihood of incarceration varied Rumbaut, R. and Miami/Ft. gitudinal study attainment, widely by parents’ country of origin, G. (2006). Lauderdale, FL looking at the incarcera- with children of Mexican or West experienc- tion, income, African parents experiencing greater es of children of marriage likelihoods of incarceration. immigrants. and parent- hood, language preference

CUNY Institute for State & Local Governance 16 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Ramey, D. M. 1999-2001 8,628 census tracts Longitudinal Study Homicide In established destination cities (cit- (2013). in 84 cities with looking at the as- and Robbery ies with well-established immigrant population over sociation between rates (National communities), increased immigra- 100,000 a) total immigrant Neighborhood tion was strongly associated with composition and Crime Study less violent crime in all four neigh- concentration and data) borhood types (White, African b) growth in the American, Latinx, and integrated). immigrant com- position on the In new destination cities (cities violent crime in- with rapidly rising immigrant com- dex. The authors munities), however, growth in looked at the in- immigration was associated on- fluence of city type ly with less violent crime in Latinx (whether the city neighborhoods. The benefits of im- historically had a migrant composition do not lead to large immigrant reductions in violent crime in the resident concentra- other 3 neighborhood types (White, tion, “destination African American, and integrated). city,” or cities where immigrant resident concentra- tion recently grew) and racial/ethnic makeup of the city.

Ruther M. 1970-2000 3,104 counties from Analysis looking at Logged homi- Findings suggested that when the (2014). the contiguous U.S. the effect of chang- cide rates from foreign-born population increased, es in foreign-born four decadal homicide rates declined, especially share of the pop- observations. in the South region of the U.S. ulation (Census) The rate is the on county homi- average num- A unit increase in the foreign-born cide rates (National ber of homicide population share in all counties was Center for Health deaths per associated with a 4.3% decrease in Statistics; National county over a homicide rates across all counties. Historical three-year pe- Geographic riod centered Information on each census System). (e.g. 1999-2001)

Immigration and Crime: A Public Policy Red Herring 17 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Sagir, M., & 2008-2014 Census tracts in Regression analysis Violent and Immigration, measured by the per- Feldmeyer, B. Cincinnati, OH looking at the rela- property crime cent foreign-born, had no impact on (2020). tionship between rates (UCR) violent, property, or overall crime the percentage of rates in the city. people who are for- eign-born (Census) There was also no indirect effect and violent and of immigration on crime using property crime. controls related to neighborhood disadvantage.

Sampson, R. 1995-2002 Individuals—2,974 Multilevel logistic Self-reported First-generation immigrants were J., Morenoff, total participant (in- regression analysis offending significantly less likely to commit J. D., & cluding 113 first- or on the association behaviors crimes than later generations (i.e., Raudenbush, second-generation between generation second-generation or higher). S. (2005). immigrants), aged status (first, sec- 18-25 years old, ond, and third or Third-generation immigrants were from the Project more generations) twice as likely as first-generation on Human and self-reported immigrants to report offending Development violent offending. behaviors. in Chicago Neighborhoods

Stacey, M., 2000-2004 All U.S. states Regression analysis Count of an- Hispanic immigration was associat- Carbone- of the relationship ti-Hispanic ed with increased anti-Hispanic hate López, K., & between increased hate crimes crimes. Rosenfeld, R. Hispanic im- (UCR) (2011). migration and However, when Hispanics consti- anti-Hispanic hate tute a larger proportion of the overall crime. population, there are reductions in hate crimes against Hispanics.

As White unemployment grows rel- ative to Hispanic unemployment, there are also reductions in anti-His- panic hate crimes.

CUNY Institute for State & Local Governance 18 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Sulkowski, 2009-2010 Individuals—2,929 Analysis looking Self-reported Immigrant youth were significantly M. L., youths (includ- at the relationship victimization more likely to report being physical- Bauman, S., ing 280 youths who between immi- (e.g., called ly hit and less likely to tell an adult Wright, S., were from families gration status and names, rumors, about abuse when compared to na- Nixon, C., that had immigrat- victimization socially exclud- tive-born youth. & Davis, S. ed to the U.S. in among youth. ed, threatened, (2014). the past two years) and physical- Immigrants reported being victim- who reported hav- ly hit) ized because of their race, religion, ing been victimized or family income at more than at least two times a twice the rate of those who were month in the Youth native-born. Voice Project

Treyger, E., 2008-2011 335 U.S. cities Differences- Crime rates No statistically significant change Chalfin, A., in-differences and arrests for in crime or arrest rates after DHS & Loeffler, C. analysis of the ef- violent, prop- verification was implemented for (2014). fect of Secure erty, and minor arrestees. Communities on crimes local crime rates There was also no reallocation of ar- and arrest behavior rests from Black suspects to White of municipal police suspects (though many of White sus- agencies. pects were also Hispanic).

Vaughn, M. Not listed Individuals— Logistic regression Violent and Immigrants were less likely to G., Salas- 41,932 (including analysis looking at nonviolent behave antisocially (violent or non- Wright, C. P., 7,320 foreign-born) the relationship be- antisocial be- violent) compared to native-born DeLisi, M., & participants in tween immigration haviors using individuals , even when controlling Maynard, B. the National status and antiso- 31 items from for risk factors (e.g., low income, R. (2014). Epidemiologic cial behavior. the antisocial men, urban areas). Survey on Alcohol personality and Related disorder mod- Native-born people were 3 to 4 times Conditions ule of the more likely to exhibit antisocial be- AUDADIS-IV haviors than immigrants from Asia, Africa, or Latin America. Non- European immigrants were the least likely to be antisocial.

Immigration and Crime: A Public Policy Red Herring 19 REFER- ENCE LEVEL OF DESCRIPTION OUTCOME(S) STUDY PERIOD ANALYSIS OF STUDY OF INTEREST FINDINGS

Wadsworth, 1990-2000 Cities with popula- Longitudinal study Robbery and When controlling for yearly trends, T. (2010) tions over 100,000 looking at the rela- homicide rates there was no relationship between tionship between (UCR) immigration and crime rates. the proportion of foreign-born res- Increases in the proportion of immi- idents and the grants overall and new immigrants proportion of for- were associated with lower homi- eign-born residents cide rates. who arrived in the Population growth among new im- last five years and migrants accounted for 9% of the the rates of robbery drop in the homicide rate and 22% of and homicide. the drop in the robbery rate.

Wu, Y., & 2002-2003 Households—4,904 Study looking at Self-reported Foreign-born Latinxs experienced Altheimer, I. respondents (includ- the relationship be- victimization higher rates of property victim- (2013). ing 567 foreign-born) tween immigration of crime ization and lower rates of violent from the Seattle status and reported victimization compared to na- Neighborhoods and victimization. tive-born Latinxs. Crime Survey

Xie, M., & 2008-2012 Individuals— 569,611 Longitudinal study Self-reporting Violence was less likely to be re- Baumer, E. P. respondents to the looking at the rela- of violent ported to police in new immigrant (2018). National Crime tionship between crime (NCVS) neighborhoods, likely due to higher Victimization changes in the barriers for immigrants rather than Survey concentration of less actual violence. foreign born and rates of report- ing violence to the police.

Zadnik, E., 2008 Individuals—1,377 Logistical Self-reporting Undocumented women were no Sabina, C., & immigrant Latina regression anal- sexual assault more likely than documented wom- Cuevas, C. A. women interviewed ysis looking at en to be the victims of violence but (2016). for the Sexual differences in re- were significantly less likely to seek Assault Among porting sexual help related to their victimization. Latinas Study assault among documented and undocumented Latina women.

Zhang, Y., 1998-2014 Alabama and 37 Synthetic control Violent and Violent crime increased significantly Palma, M. A., comparison states analysis look- property crime in Alabama following HB 56. & Xu, Z. P. ing at the effect rates (UCR) (2016). of Alabama HB There was no change in property 56, “the strictest crime after HB 56. anti-illegal im- migration law in the U.S.,” on vio- lent and property crime.

CUNY Institute for State & Local Governance 20