Secure Communities by the Numbers, Revisited (Part 2 of 3)

Analyzing the Analysis


Analyzing the Analysis

Download a pdf of this Memorandum

Read Part 1 - Read Part 3


Jessica M. Vaughan is Director of Policy Studies at CIS. W.D. Reasoner (a pseudonym) is a retired government employee with many years of experience in immigration administration, law enforcement, and national security matters.


Summary

This report is the second in a series examining outcomes of the Immigration and Customs Enforcement (ICE) agency's Secure Communities program and how those outcomes have been misleadingly described by one widely circulated paper, "Secure Communities by the Numbers: An Analysis of Demographics and Due Process,"1 from the Warren Institute on Law and Social Policy at the University of California, Berkeley, Law School, and then uncritically re-told by major news media outlets. The Warren Institute report and our reports are based on the same database of actual case histories provided by ICE in response to a Freedom of Information Act Request.

Former ICE Official Dan Cadman Explains
the Secure Communities Program:
View the Full Interview

Our first review focused on assertions relating to U.S. citizens, whom the Warren report erroneously alleged had been taken into custody as the result of the Secure Communities (SC) initiative, and placed into civil removal (deportation) proceedings.2

This, our second review, focuses on the set of assertions relating to alleged discrepancies in the number of Latinos taken into custody by Secure Communities, versus national averages, which the Warren report cites as potential evidence of racial or ethnic profiling. Our findings:

  • The Warren Institute report asserts that 93 percent of the individuals apprehended were "Latino". Yet ICE databases do not contain information relating to a person's race or ethnicity. Making judgments of race and ethnicity based solely on nationality, as the Warren Institute report authors did, is speculative and prone to error, for reasons we explain in the body of this report.

  • The Warren Institute report describes, without differentiation, all of the apprehended "Latino" individuals contained in the dataset as "undocumented". By our calculation, slightly more than 12 percent of those arrested were in fact documented, and included nonimmigrants, immigrants, border crossing card holders, and both conditional and permanent resident aliens. As we observe in more detail below, even documented aliens are subject to removal from the United States if they violate the conditions of their stay — particularly if they do so by committing crimes, which is the focus of the Secure Communities initiative. This distinction is important and should have been noted by the authors of the report.

  • The Warren Institute authors compared their calculation that 93 percent of arrested aliens in the dataset were Latino to the Latino share of the national illegal alien population of 77 percent, presenting this as evidence of ethnic bias in the operation of Secure Communities. This is a mistake because 81 percent of cases in their database are drawn from just three southwest border states (Arizona, Texas, and California). These states, according to DHS, have only 45 percent of the illegal population. More importantly, illegal aliens in these three states are disproportionately Latino. For example, it is estimated that 94 percent of illegal aliens in Arizona are from Mexico, whereas 62 percent of all illegal aliens are estimated to be Mexican.

  • The presumed ethnic profile of the cases in this database (based on country of nationality) very closely matches the ethnic profile of the population of criminal aliens nationwide and also in the states where most of the SC arrests took place. According to a variety of government and independent sources, the population of criminal aliens nationwide and in Texas, California, and Arizona, where most of the SC arrests in the database took place, is approximately 90 percent Latino, which is nearly identical to the percentage of Latino detainees in the database (93 percent). This undercuts the claims that Secure Communities results in a disproportionate number of Latino arrests.

  • The Warren Institute report suggests that statistics gleaned from the database bolster allegations that the Secure Communities program is having a negative effect on community policing and may be masking illegal racial profiling practices. Neither allegation is supported by information in the database, nor by reputable academic or government studies.

  • Over the course of our analysis it became evident that many of the cases in this database do not relate to the Secure Communities program at all. About 6 percent of the cases were attributable to ICE's sister agency, Customs and Border Protection, and another large share of arrests took place in jurisdictions where Secure Communities had not yet been activated. This could be due to errors on ICE's part in fulfilling the data request as required. In our judgment, the Warren Institute authors should have detected the data discrepancies and noted them in their report.

  • We regretfully conclude that this data source is of no value in assessing the Secure Communities program. In our view, ICE has squandered an opportunity to exhibit transparency and shed light on the workings of one of its most important enforcement programs.

Introduction and Recap

In October of 2011, the Earl Warren Institute at the University of California, Berkeley, Law School and the Benjamin N. Cardozo School of Law jointly published a report, "Secure Communities by the Numbers: An Analysis of Demographics and Due Process", which is available at http://www.law.berkeley.edu/files/Secure_Communities_by_the_Numbers.pdf. The report (referred to, from here onward, simply as "SCBTN") focuses on the Secure Communities program operated by the Immigration and Customs Enforcement (ICE) bureau within the Department of Homeland Security (DHS). The Secure Communities initiative uses electronic systems to compare fingerprints of individuals taken into custody by state and local police against a DHS fingerprint repository to identify aliens arrested for crimes who may be subject to removal (deportation) from the United States.

SCBTN made some disturbing findings and assertions, among which were the following:

  • Arrests of United States Citizens. "Approximately 3,600 United States citizens have been arrested by ICE through the Secure Communities program."

  • Racial/Ethnic Profiling. "Our analysis … raises serious concerns about the level of screening and potential targeting of certain social groups … . Latinos comprise 93 percent of individuals arrested through Secure Communities though they only comprise 77 percent of the undocumented population in the United States … . Community and advocacy groups have also asserted that Secure Communities is, in some jurisdictions, masking local law enforcement agencies' practice of racial profiling."

  • Due Process and the Operation of Law. "Only 24 percent of individuals arrested through Secure Communities and who had immigration hearings had an attorney compared to 41 percent of all immigration court respondents who have counsel … . Only 2 percent of non-citizens arrested through Secure Communities are granted relief from deportation by an immigration judge as compared to 14 percent of all immigration court respondents who are granted relief."

We decided to review the analysis and findings of the report, to determine whether they were well founded. We obtained from ICE officials the same dataset that they had given to the report's authors so that we could examine it for ourselves. A read-only copy of the dataset that was provided to us can be found at http://www.cis.org/sites/cis.org/files/articles/2011/sc-by-numbers-datas... , for those who wish to examine it themselves.

We determined that the SCBTN report is flawed in several areas, and decided that publishing our own findings was important in the public interest, because a number of special interest groups, open borders advocates, journalists, and members of Congress have quoted the report3 — including those portions that are most erroneous or suspect — in interviews and articles with media outlets, and at hearings.

We also concluded that ICE materially contributed to acceptance of the report's erroneous findings through its silence and failure to routinely provide timely and accurate statistical information.

Because of our concerns, we decided to prepare a series of papers to discuss the report's analysis and findings, divided into three broad areas, consistent with what we perceive of as the main themes addressed in SCTBN:

  • arrests of United States citizens;

  • allegations of racial or ethnic bias; and

  • due process and operation of the rule of law, versus perceived flaws in the program.

We published our first paper, regarding the arrests of United States citizens, in December 2011. In it, we specifically found that SCBTN's authors did not fully understand the data, resulting in significant errors of review and interpretation; that their methodology in arriving at certain conclusions was unsound; and, consequently, that they engaged in overbroad, insupportable assertions based on their philosophical views rather than the facts.

This is our second paper, focusing specifically on the report's assertions with regard to racial or ethnic bias in the Secure Communities program. We have again found that troubling methodology, unsupportable assertions, and unsound analogies were used by the SCBTN authors in attempting to make the case that ICE's Secure Communities program is flawed.

Apprehension Rates of Latinos and Ethnic Profiling

Among the assertions made in the SCBTN report is that, "Our analysis provides a fuller picture of the population being processed through Secure Communities and raises serious concerns about the level of screening and potential targeting of certain social groups … ."4 Specifically, it states that "Latinos comprise 93 percent of individuals arrested through Secure Communities though they only comprise 77 percent of the undocumented population in the United States."5

The authors go on to state, "Community and advocacy groups have also asserted that Secure Communities is, in some jurisdictions, masking local law enforcement agencies' practice of racial profiling",6 and therefore recommend that "a thorough investigation [be conducted] into the potential racial profiling of Latinos as a result of Secure Communities and implement safeguards to protect against such abuses … ."7

A critical analysis of the claims in the report requires inquiry into four discrete areas:

  1. Is the SCBTN assertion, that 93 percent of the apprehended aliens in the sample data were Latino, supportable?

  2. Were they in fact "undocumented"?

  3. Is comparison of the percentage of apprehended Latinos in this dataset to the national average of 77 percent relevant or accurate?

  4. Is there evidence of ethnic profiling in the arrests by state and local police?

Having reviewed the dataset on which SCBTN was based, we agree neither with its statistical accuracy, its conclusions, nor its recommendations.

And, finally, we asked two more questions that flowed from our examination of the questions above.

  1. Do the statistics mean that Latino immigrants are more likely to be criminals than others?

  2. Does Secure Communities damage police relations with immigrants?

Following is our analysis in full.

1. Is the SCBTN Assertion, that 93 Percent of the Apprehended Aliens in the Sample Data Were Latino, Supportable?

Our analysis: It is not supportable using the available data.

It is important to note right from the start that no ICE arrest reports or forms provide data fields for officers to insert observations about arrestees' ethnicity, nor for the arrestees to self-identify their ethnicity. The authors of the SCBTN report acknowledge this, though they state it in a way that obfuscates its import.8 So how did they arrive at that percentage? Absent hard data, they were left to consider "Latino" ethnicity based on country of citizenship — a minefield if ever there was one, insofar as it generally assumes that all citizens of Latin American countries are, in fact, Latino.

Out of curiosity, we attempted to replicate their analysis, although we reject the initial premise. By removing duplicates, and then eliminating U.S. citizens in the dataset, we were then forced to decide which countries were sufficiently "Latino" to count.9 Using the crude hammer of "country of citizenship" as a basis for presumptions of ethnicity, we obtained a figure of 92.6 percent, which is essentially identical. Why then do we believe that the report's assertion is insupportable?

Because it generally assumes that all persons who are citizens of, for instance, Guatemala are Latino, notwithstanding that 41 percent of the population are non-mestizo Amerindians who, if they speak Spanish, speak it as a second language, not their mother tongue, and that they may not think of themselves as Latino given their Native American heritage. Or that all Bolivians, Ecuadorians, or Peruvians are Latinos despite the fact that in those countries there are large percentages of individuals of almost exclusively Quechua or other Amerindian descent, as well as significant populations of Asians. Or that all Hondurans are Latino, despite the fact that there are pockets of the country in which there are individuals of Middle Eastern descent. Or even that arrestees from Mexico, by far the largest contributor to the illegal alien population of the United States, are all Latino despite the fact that 30 percent of the Mexican population consists of non-mestizo Amerindians.10

For all of these reasons, it is statistically suspect to gauge individuals' "Latino-ness" based on countries of citizenship. Not every national of a so-called Latin American country is Latino, due to substantial indigenous, non-mestizo populations; importation of large numbers of African slaves to the New World in past centuries; and more recent migratory trends of non-Hispanic populations from Asia, the East Indies, the Middle East, and other areas into the Latin American regions.

In reviewing the work of several other researchers, including the Pew Hispanic Center, we found none willing to venture into the wilderness of ethnic breakdowns of the illegal alien population of the United States based solely on countries of origin.11 The Census Bureau figures on ethnicity, for example, are based on self-identification.

There is something repugnant about presuming a kind of monolithic racial or ethnic aspect to human beings from so many different nations, which were and are themselves melting pots of cultures and places. It is a particularly curious approach to take for organizations that represent themselves as watchdogs of human rights and diversity.

2. Were the Presumed Latinos Who Were Apprehended All "Undocumented"?

Our analysis: No. The SCBTN authors appear to have posited that all of the presumed-Latinos in the dataset were "undocumented", but they were not. We determined that slightly more than 12 percent of those individuals had some form of status — parole documents, border crossing cards, nonimmigrant visas, immigrant visas, and conditional or permanent resident cards.12

Here is how we undertook our calculations: First, we examined the information for each case in the tab of the original dataset identified as Item #15 (this tab shows the country of citizenship of the aliens apprehended). We then eliminated all duplicate records of arrest for the same individuals. Next, as we did previously, we eliminated all cases except those presumed to be "Latinos" (see End Note 9 for a list of the nationalities presumed to be Latino). This left us with a residual pool of 461 presumed Latino aliens in this subset of data. Finally, we cross-referenced those cases against the status-at-entry tab identified as Item #24. The result is reflected in Table 1.

There are two important caveats readers should note: 1) there was a discrepancy between the number of cases found in Tabs 15 and 24. Ideally, the cases should have matched exactly so that we could correlate all data. But this was one of many flaws in the dataset as provided by ICE. 2) For this reason, where cases appeared on Tab 15 but not on Tab 24; or where the entry status field for an individual in the remaining subset was left blank; or the status at entry information was unclear (e.g. "other"), we gave the SCBTN authors the benefit of the doubt and counted the individual as "undocumented". But even with this generous assessment, we still determined that 12.1 percent of the Latino arrestees (56 individuals) were in fact documented.

Why did we undertake this particular examination? Because words, like statistics, are important. They lose their meaning when used imprecisely or inaccurately. We dislike the phrase "undocumented" because it is imprecise. Not every alien who is subject to ICE apprehension and removal is without documents. Documented aliens — even lawful permanent residents — become removable (deportable) for certain offenses; for instance: engaging in fraud, participating in alien smuggling schemes, or committing a crime.13

This latter basis for deportability is particularly relevant to the study, since the whole point behind the Secure Communities fingerprint matching initiative is to identify aliens who have been taken into custody by local police and charged with crimes — even, when appropriate, resident aliens. Thus, the indiscriminate imputation of the word "undocumented" to all of the presumptive Latinos is not only imprecise, it is downright inaccurate, and should not have been used.

However, we also want to observe that in our prior paper reviewing the SCBTN report (Part 1, relating to the report's assertions regarding United States citizens), we expressed our concerns over the data integrity problems we confronted when dealing with the ICE dataset. We again have been confronted with those problems in this paper, as is obvious by the limitation on our attempt to differentiate "documented" from "undocumented" individuals apprehended as the result of Secure Communities, due to incomplete information fields and case data lists that did not match in size and number.

We recommended in Part 1 that ICE should routinely validate and publicize its Secure Communities data and, when necessary, provide explanatory comments or insert appropriate, carefully worded caveats about the use and limitations of the data. We also suggested that if ICE is incapable of doing so, then they should cede that responsibility to the DHS Office of Immigration Statistics, which consistently does a credible job of publishing yearly statistical information of interest to researchers all across the philosophical spectrum. We reiterate the importance of those recommendations here.

3. Is Comparison of the Percentage of Apprehended "Latinos" in this Dataset to the National Average of 77 Percent Relevant?

Our analysis: It is not.

First of all, the SCBTN authors used an outdated estimate for the origins of the national illegal alien population. The source they cite, a 2006 study by Jeffrey Passel of the Pew Hispanic Center, is a reliable one, but it is not the most recent study available. In February 2011, the Pew Hispanic Center published another study by Passel that includes an estimate of the countries of origin of illegal aliens in 2010.14 That study estimates that the share of the illegal population that is Latino to be about 80 percent, which is slightly higher than the 2006 estimate of 77 percent. The 2010 estimate of 80 percent is also closer to the time period covered by the ICE database.

Regardless of the source, the comparison of the individuals in this limited database to the national illegal population is an invidious apples-and-oranges comparison, for a number of reasons.

To remind readers, in the first place we believe that estimating the percentage of Latinos among the dataset studied based solely on country of citizenship is insupportable. The same logic was used by the SCBTN authors to derive the national average of 77 percent; we believe that figure, too, is insupportable for the same reason — because it does not take into account the actual demographics of Latin American countries. It simply glosses over the whole complex variety of racial and ethnic origins of the peoples of those countries and, by statistical sleight of hand, makes them "Latino".

But if we put that signal fact aside and accept the SCBTN Latino percentage assumptions, solely for the sake of exposition, the comparison still serves no legitimate purpose, for this simple reason: It is inappropriate to apply a national average to a program that has not been activated nationally.15

Readers should be made aware that, even now, Secure Communities is not fully active nationally. At the time the snapshot that this dataset represents was taken (October 1, 2008, to January 31, 2010), the number of states and jurisdictions participating was even more limited. However, ICE had focused on fully activating the Secure Communities program in the southern and southwestern border states as a way of complementing other ongoing DHS and Justice Department initiatives designed to mitigate the effects of cross-border crime and violence spilling over from Mexico. In those parts of the country, there is a considerably larger Hispanic population, both legal and illegal, than in many other parts of the United States, and it would stand to reason therefore that percentages of "Latino" apprehensions would be higher.

With that in mind, we decided to look more closely at the data. Once again we are obliged to explain our methodology. As before, we derived this table from information gleaned in two tabs: the one labeled Item #19 (Location of Arrest); and the one we used previously, labeled Item #15 (Citizenship Country), but modified to reflect only the presumed "Latino" portion of those apprehended. And, as before, we noted that there was an inexplicable discrepancy in the dataset between the number of cases found in Tab 15 versus the number found in Tab 19, so where we could not discern the place of arrest, we were obliged to set aside the record. This left us with a pool of 438 cases. Table 2 shows the complete tabulation by state.

Here is what we discovered in our examination:

  • The dataset consisted of apprehensions made only in 14 states — not all 50, nor even the 48 continental states, as one would be led to believe from the SCBTN report.

  • In eight of the 14 states, there were five or fewer arrests.

  • Only three states — Arizona, California, and Texas — contributed nearly 80 percent of all the arrests in the dataset. This is not unexpected: all three are southwestern states that share borders with Mexico. All have a larger than average Latino population, and non-citizen Latino population in particular.

These three states, according to DHS, have only 45 percent of the illegal population.16 More importantly, illegal aliens in these three states are disproportionately Latino. For example, the Pew Hispanic Center estimates that 94 percent of illegal aliens in Arizona are from Mexico, whereas 62 percent of all illegal aliens are estimated to be Mexican.17 It is likely that other Latino groups would add perhaps two to four percentage points to those figures.

Thus, we find that the apprehension rates of putative "Latinos" in the dataset are entirely consistent with the regions in which the apprehensions took place.

4. Is There Evidence of Ethnic Profiling in Arrests by State and Local Police?

Our analysis: We found none.

In considering the possibility of ethnic profiling, we asked this question: How does the imputed ethnicity of the individuals in the database compare with the population in the states and localities where most of the arrests took place?

The SCBTN authors concluded that Latinos were disproportionately affected by Secure Communities because the number of people identified for deportation through SC in the database who were from "Latino" countries (93 percent) was disproportionate to the number of illegal aliens nationwide who are from "Latino" countries (77 percent). They suggest that this statistic supports the claims of certain advocacy groups that Secure Communities serves to mask local law enforcement agencies' practices of racial profiling. This finding has been widely circulated and is frequently mentioned by Secure Communities skeptics as evidence that the program must be ended or drastically scaled back. In a recent oversight hearing before the House immigration subcommittee, several Democratic members of the subcommittee and one witness cited this finding as a major reason for their concerns about the program.18

But there are a number of problems with the SCBTN analysis, and especially with the allegations about racial profiling. We have already noted the data integrity and selection bias issues that we believe contaminated the SCBTN analysis from the start.19 Nevertheless, because the statistics generated by the SCBTN study have taken on a life of their own, we feel it is important to examine them, and to offer some context in which to evaluate them.

As we have stated, it is inappropriate to compare the ethnicity of the individuals in the SCBTN database to the national illegal population as the SCBTN authors did, because the Secure Communities program was not operating nationally during the time period under examination. And as we have noted, the majority (80 percent) of arrests occurred in Texas, California, and Arizona. Two of these states, California and Texas, have by far the largest Latino populations in the country, together accounting for about one-half of the nation's entire Latino population.

Further, more than half of the arrests (239, or 51 percent) took place in four of the five U.S. counties that have the largest Latino populations (Los Angeles County, Calif.; Harris County, Texas; Miami-Dade County, Fla.; and Maricopa County, Ariz.).20 Considering that the arrests took place in areas of the country that have larger than average Latino populations and, more importantly, that the large majority of illegal aliens in the entire country are citizens of "Latino" countries, it is not surprising that most of the arrests in the SC database were individuals who are citizens of "Latino" countries.

It would be more appropriate to make a comparison of ethnicity to the illegal populations within the states or metropolitan areas where most of the arrests took place. However, we could find no localized estimates on the ethnicity or country of citizenship of the illegal alien population.

Besides, it is important to remember that the Secure Communities program does not operate within or touch the entire U.S. population, nor even the entire non-citizen or illegal alien population. The program touches only those individuals who are arrested or booked into custody by a state or local law enforcement agency. Furthermore, the purpose of Secure Communities is to identify those individuals in custody who are potentially removable. That would be non-citizens — mainly illegal aliens, but also lawful permanent residents and others with legal status who commit crimes that are serious and/or numerous enough to render them deportable.

Merely stating, as the SCBTN authors do, that the proportion of Latinos in the SC case load exceeds the proportion of Latinos in the national illegal alien population does not demonstrate that there is racial profiling occurring. To make the case that Secure Communities is disproportionately affecting Latinos, or that the program is masking racial profiling, the authors would have to show that ICE is disproportionately selecting Latinos for removal from among all the apparent non-citizens presented to the agency as a result of the interoperability, and/or that local law enforcement agencies are arresting a disproportionate number of Latinos, either out of deliberate general bias against Latinos or for the specific purpose of triggering the immigration status check.

Neither of those theories was tested by the SCBTN researchers. First, the SCBTN researchers studied only those records of individuals who were selected by ICE for removal processing (502 out of the 1,650 total records of IDENT hits). They did not compare this population to the total population of non-citizens flagged by Secure Communities interoperability, and thus cannot draw any conclusion about profiling on ICE's part. Nor did the researchers examine how the population in this database compares to the criminal or criminal alien population nationwide or in the states where SC was operating.

Because the SCBTN authors had not done so, we decided to test the allegation of disproportionate Latino arrests using data from a variety of public sources. We find that the presumed ethnic profile of the SC caseload in this database (based on country of citizenship) closely matches the profile of the population of criminal aliens nationwide and also in the states where most of the SC arrests took place. We found that the population of criminal aliens nationwide and in Texas, California, and Arizona, where most of the SC arrests in the database took place, consists very predominantly of citizens of "Latino" countries. Our conclusion is based on a review of statistics and reports from a variety of state and federal government agencies, some obtained through Freedom of Information requests, and also independent reports. See Table 3.

Since the nationality of the individuals identified through Secure Communities is consistent with the nationality of aliens arrested nationwide and especially in the states where it is operating, we find no support for the allegation that this program is causing or masking local practices of racial profiling. This conclusion is not to deny that racial profiling might occur; we simply don't know based on the data available. But we can state absolutely that it has not been established by the simplistic analysis conducted by the SCBTN researchers.

Nationality of SCBTN Caseload Compared to DHS Caseload. Let's examine these cases using one final point of analysis. We compared the nationality rankings of apprehended aliens in the SCBTN database with the nationality rankings of aliens apprehended and removed by DHS officers in 2010, as reported by the Office of Immigration Statistics in its annual Immigration Enforcement Actions report.21

This comparison, illustrated in Table 4, reveals a very close parallel. We find that the case load in the SCBTN database is general reflective of, and not significantly deviant from, the overall DHS workload as expressed in both apprehensions (primarily CBP workload) and removals (CBP and ICE workload). This suggests to us that the condemnation of the Secure Communities program in particular by the SCBTN authors is baseless and without merit.

What's more, this comparison shows that the case load in this database is actually more diverse than the national DHS case load. Table 4 shows that the presumed Latino share of the SCBTN case load, at 93 percent, is lower than the presumed Latino share of the overall DHS case load, whether measured by apprehensions or removals.

This finding casts some doubt on the allegation by the SCBTN authors that Secure Communities is masking racial profiling by local law enforcement agencies, or encouraging them to make pre-textual arrests in the hopes of discovering illegal aliens. If such illegal profiling or abuse were occurring, then one would expect that the SCBTN database would have a higher proportion of Latino arrests, not lower, as we found.

As we have discussed, the cases in the database came primarily from states along the U.S.-Mexico border with larger-than-average Latino populations. As the Secure Communities program interoperability is slowly deployed nationwide, then the ethnic profile of the case load most likely will change somewhat, and become even less "Latino", because it will better reflect the ethnic profile of the national illegal alien population.

This suggests that, if there is concern about the degree of diversity in the population that is subject to immigration law enforcement, then one effective way to address that concern would be to accelerate the deployment of Secure Communities nationwide to parts of the country that have more diverse and proportionately less-Latino illegal alien populations — places such as New York, Illinois, and Massachusetts. Ironically, the critics of Secure Communities, including authors of the SCBTN report, have worked hard to obstruct that nationwide deployment.

Conclusion. We have examined the Latino population through three reasonable and statistically sound prisms: geography (locality of arrest); criminality (relevant prison populations); and immigration enforcement apprehensions (which are primarily undertaken on the southwestern border of the United States). Nothing supports the assertion that there is evidence of bias or profiling. In fact, in our view, the allegation of ethnic profiling is simply the top floor in a deck of cards. We have pulled away the foundation and lower floors, and there is nothing left on which the SCBTN authors could reasonably have based such an allegation.

5. Do These Statistics Mean that Latino Immigrants Are More Likely to Be Criminals than Others?

No. We want to emphasize that these statistics on the citizenship/ethnicity of the criminal alien population cannot be interpreted or used to draw conclusions about the level of criminality in the immigrant population as a whole, or even the illegal immigrant population as a whole. While our prior research has shown that there are many legitimate public safety reasons for local law enforcement agencies to pay attention to the immigration status of offenders, we also have found that there is no evidence that non-citizens commit crimes at higher or lower rates than other groups. There are many other factors that influence these statistics, including age, socio-economic circumstances, and more. For further discussion of these issues, see the CIS publications Immigration and Crime: Assessing a Conflicted Issue and Taking Back the Streets: ICE and Local Law Enforcement Target Immigrant Gangs.22

6. Does Secure Communities Damage Police Relations with Immigrants?

Not necessarily. Contrary to assertions made in the SCBTN report, there is little evidence that cooperative enforcement programs such as Secure Communities have a negative impact on police relations with immigrants in their community. This is especially true where police departments and ICE have robust community policing and outreach programs that explain the goals of the immigration enforcement programs.

Very little substantive research on this question has been done, but the few studies that have examined the relationship between community policing and immigration law enforcement have found that while there may be some reluctance to report certain crimes among immigrants, this is mostly due to language barriers and cultural differences, not a fear of being targeted for deportation.23 According to the Bureau of Justice Statistics surveys on crime victimization and reporting, while only about 50 percent of all crimes are ever reported to police, Latino individuals are just as likely to report them as non-Latinos.24 In addition, the statistics available both nationally and locally indicate that crime reporting rates do not decrease disproportionately when local law enforcement agencies assist with immigration enforcement.25

In our observation, the negative perceptions about immigration enforcement and immigrant community relationships with police are generated mainly by advocacy groups that oppose immigration law enforcement, often in response to a perceived increase in enforcement such as the implementation of Secure Communities, and are not typically based on the experiences or attitudes of immigrants who live in the community, nor on empirical or statistical analysis.

Data Problems

Over the course of this analysis, we encountered numerous flaws in the dataset. We believe these flaws were so obvious that they should have been noticed by the SCBTN authors and acknowledged in their report. We further believe that these defects so undercut the integrity of the data that they make it impossible to draw any conclusions about the Secure Communities program from this dataset.

For example, when analyzing the locations of arrest, we noticed that 28 of the individuals (representing 6 percent of the total) were coded as having been apprehended at ports of entry or Border Patrol stations, presumably by Customs and Border Patrol (CBP) Officers. Most are not listed as having been booked into a detention facility. Clearly, these individuals were not identified through the Secure Communities program interoperability.

One singular case (a "non-Latino", using the SCBTN authors' rough methodology, though in fact he or she may have been a Hispanic Canadian; we have no way to know) was coded as having been encountered in Montreal, Canada. This individual was probably a denied applicant seeking to board a flight en route to the United States at the U.S. pre-inspection facility in Montreal International Airport. Although the denied applicant may certainly have been an IDENT match when the CBP inspector took his or her fingerprint prior to deciding whether to permit boarding — the match may in fact have been the reason for the denial — this clearly could not have been a Secure Communities arrest (U.S. inspectors have no arrest authority in Canada), and should not have appeared on the dataset.

We further noticed that many of the locations of arrest were in areas that had not yet had Secure Communities activated. For example, 16 percent of the California arrests were in non-SC locations. Other obviously non-SC locations were New York City (which has refused to participate), JFK International Airport on Long Island, and a number of state correctional facilities. These prisons generally do not submit the type of fingerprint transactions that would be automatically cross-matched against the DHS IDENT database under Secure Communities interoperability principles.

We asked ICE to explain why these non-Secure Communities cases were included in the database. According to the court stipulation, ICE was ordered to provide "a list of all individuals identified as 'IDENT-Matches' by the Secure Communities program between the dates of October 1, 2008, and January 31, 2010". The SCBTN authors describe the database as having been extracted from "all IDENT matches (Secure Communities fingerprint queries that resulted in 'hits' in the Department of Homeland Security's databases)".

ICE officials familiar with the transaction that resulted in creation of the database, speaking off the record, told us that ICE believed that it was to provide records of arrests from all ICE programs, including Secure Communities, the Criminal Alien Program, the Investigations division, and any other ICE program or activity. They said that all DHS agencies were required to provide records of arrests for encounters that matched the individuals selected in the random sample. They suggested, for example, that if an individual were apprehended by the Border Patrol and found to be the subject of a warrant for a state crime, that individual would have been turned over to the local law enforcement agency, booked into a local facility, screened through Secure Communities, and thus might appear in the database with all the relevant records. But that explanation falls short.

While we did find one or two records that possibly indicated such a scenario, this could not explain the vast majority of the obviously non-Secure Communities records. Most of the cases mentioned above as likely CBP arrests have no U.S. detention facility or other local arrest or charge listed in the record. It is impossible that the random sample was drawn from a database of all of the DHS agencies' arrest records for that time period; if so, the database would contain a much higher proportion of CBP cases, because that agency makes far more apprehensions overall than ICE does.

Instead, judging from the locations of arrest and the detention facilities listed, we believe that this database includes primarily records of individuals arrested by all of ICE's Enforcement and Removal Operations programs, including Secure Communities, the Criminal Alien Program, the Institutional Removal Program, 287(g), the Fugitive Operations program, and routine referrals from local law enforcement agencies. We do not understand the presence of CBP arrest records; most are clearly not linked to a local arrest.

It is our understanding that ICE data systems include tools to enable the identification of cases according to the program that is responsible for the arrest. A number of published ICE reports include information classified in that way. We have no explanation for why ICE did not provide data that truly related solely to the Secure Communities program, when it appears to have the ability to easily do so. In failing to provide what the requestors asked for, and what the court ordered, we believe that ICE has misled the researchers and, equally importantly, squandered an opportunity to exhibit transparency and provide useful and interesting information to the public about a program that has become very controversial.

Next Report

In our next and final report in this series, we will examine the allegations made by the SCBTN authors regarding alleged disrespect or disregard of aliens' rights, due process, and the operation of immigration law.


End Notes

1 Aarti Kohli, Peter L. Markowitz, and Lisa Chavez, "Secure Communities by the Numbers: An Analysis of Demograpjics and Due Process", the Chief Justice Earl Warren Institute on Law and Social Policy, Berkeley, Calif., October 2011, http://www.law.berkeley.edu/files/Secure_Communities_by_the_Numbers.pdf.

2 See Reasoner and Vaughan, "Secure Communities by the Numbers, Revisited", Center for Immigration Studies, December 2011, http://cis.org/SC-by-the-numbers-critique-part1.

3 See, for instance, the statements made by Felipe Matos of the organization Presente.org during a television interview with Fox News on October 24, 2011, http://cis.org/TVInterviews/Camarota-FOX102411-SecureCommunities.

4 "Secure Communities By the Numbers", p.13, Conclusions.

5 "Secure Communities By the Numbers", p. 2, Key Findings, and p. 5, Ethnicity.

6 "Secure Communities By the Numbers", p. 6, Ethnicity, continued from the prior page of the report.

7 "Secure Communities By the Numbers", See p. 23, Conclusions recommendations box).

8 "Secure Communities by the Numbers", p. 5, Demographic Data, "There is no definitive dataset on the characteristics of individuals who are placed in deportation proceedings in the United States. To see whether any patterns emerge in this report, we compared the age, gender and country of origin of people in our sample population with existing datasets on noncitizens and on unauthorized immigrants."

9 The countries were Bolivia, Colombia, Cuba, Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Peru, and Venezuela.

10 See the Central Intelligence Agency's World Factbook at https://www.cia.gov/library/publications/the-world-factbook/. Look specifically in the "People and Society" section of a particular country for racial and ethnic breakdowns.

11 See, for instance, the Pew Hispanic Center's statistical data on national origins of the illegal alien population of the United States, at http://pewhispanic.org/unauthorized-immigration/.

12 Although we recognize that parole is not, technically, an admission under the immigration laws, we believe it is appropriate to deem those aliens who were paroled into the United States by federal officers as "documented" — it is particularly relevant, for instance, to Cuban "Latinos" since under the law Cuban nationals may apply for adjustment to permanent resident alien status one year after having been paroled into the United States.

13 In fact, of the 56 documented aliens, 22 of them — almost 40 percent — were identified as lawful permanent resident aliens, conditional resident aliens, or immigrants. (Note that immigration enforcement officers of DHS do not use the term immigrant loosely, as in, "illegal immigrant." Rather, in this context, it is one who possesses an immigrant visa.)

14 Jeffrey Passel and D'Vera Cohn, "Unauthorized Immigrant Population: National and State Trends, 2010", Pew Hispanic Center, February 1, 2011, http://www.pewhispanic.org/2011/02/01/unauthorized-immigrant-population-....

15 The authors knew this because they make tangential reference to where, how, and when Secure Communities activation has taken place on the very first page of the report (see the second paragraph of the Introduction). But as with other important information and data, they chose not to apply this knowledge where it was most appropriate, opting instead to make an unfounded comparison that appears to better fit their philosophical preconceptions.

16 Michael Hoefer, Nancy Rytina, and Bryan C. Baker, "Estimates of the Unauthorized Immigrant Population Residing in the United States: January 2010", Department of Homeland Security, Office of Immigration Statistics, February 2011, http://www.dhs.gov/files/statistics/immigration.shtm.

17 "Hispanics and Arizona's New Immigration Law", Pew Hispanic Center Fact Sheet, April 29, 2010, http://pewhispanic.org/files/factsheets/68.pdf.

18 See U.S. House Subcommittee on Immigration Policy and Enforcement hearing, "Is Secure Communities Keeping Our Communities Secure?" November 30, 2011, http://judiciary.house.gov/hearings/hear_11302011.html.

19 See Reasoner and Vaughan, "Secure Communities by the Numbers, Revisited", Center for Immigration Studies, December 2011, http://cis.org/SC-by-the-numbers-critique-part1.

20 No arrests in this database occurred in Cook County, Ill., which is one of the top five Hispanic counties. As a matter of policy, Cook County actively obstructs ICE from taking custody of criminal aliens in its facilities.

21 Department of Homeland Security, Office of Immigration Statistics, Immigration Enforcement Actions 2010: Annual Report, June 2010, http://www.dhs.gov/xlibrary/assets/statistics/publications/enforcement-a.... (See, specifically, the subsection "Nationality of Apprehended Aliens" and Table 1, on page 3 of this report.)

22 Steven A. Camarota and Jessica M. Vaughan, Immigration and Crime: Assessing a Conflicted Issue, Center for Immigration Studies, November 2009, http://cis.org/ImmigrantCrime ; and Jessica M. Vaughan and Jon Feere, Taking Back the Streets: ICE and Local Law Enforcement Target Immigrant Gangs, September 2008, http://cis.org/ImmigrantGangs.

23 Robert C. Davis and Edna Erez, "Immigrant Populations as Victims: Toward a Multicultural Criminal Justice System", National Institute of Justice: Research in Brief, May 1998, and Davis, Erez, and Avitable, "Access to Justice for Immigrants Who Are Victimized: The Perspective of Police and Prosecutors", Criminal Justice Policy Review 12:3, September 2001.

24 See Criminal Victimization Statistics, http://www.ojp.usdoj.gov/bjs.

25 See, for example, University of Virginia Center for Survey Research, Evaluation Study of Prince William County Police Illegal Immigration Enforcement Policy: Final Report 2010, http://www.pwcgov.org/government/bocs/documents/13188.pdf and Jessica M. Vaughan and James R. Edwards, Jr., The 287(g) Program: Protecting Home Towns and Homeland, Center for Immigration Studies, October 2009, pp. 18-20, http://www.cis.org/287greport.