April 15, 2009

V. Analyzing the ICE Dataset

Aggregate Data

Total number of persons deported on criminal grounds 1997-2007

The dataset Human Rights Watch acquired from the Immigration and Customs Enforcement agency of the US Department of Homeland Security in August 2008 contained information on 897,099 people who were deported on criminal grounds between April 1, 1997, and August 1, 2007.[34]

In those 10 years and 4 months, deportations occurred each and every day, representing 3,775 individual days in which non-citizens were loaded onto buses or planes and removed from the United States. Figure 1 below reveals the number of persons deported during each calendar year (including partial years) for which we have data. As the figure shows, there has been an almost consistent annual increase in the number of deportations for criminal conduct. In 1998, the first year for which we have complete data, 72,482 non-citizens were deported on criminal grounds. Whether due to stepped-up enforcement,[35] or simply reflecting annual non-citizen population growth, by 2006 the number had increased 42 percent from its 1998 level, to 103,163 deportees.


Figure 1: Total Deportations on Criminal Grounds Based on ICE Dataset

These aggregate numbers suggest significant data management problems at ICE. In Table 2 below we compare the dataset that ICE supplied to Human Rights Watch with data published by the DHS Office of Immigration Statistics annually in its Yearbook of Immigration Statistics.[36] The table only includes years for which we have complete data. There is a significant discrepancy between these two sources regarding the number of non-citizens deported from the US on criminal grounds. Assuming the ICE dataset provided privately to Human Rights Watch is correct, the public DHS Yearbooks of Immigration Statistics are failing to account for between 5.6 percent and 14.3 percent of deportees.

Table 2: Comparing DHS Dataset Provided to HRW with Agency Publications [37]


Number Deported on Criminal Grounds in DHS Publications

Number Deported on Criminal Grounds in ICE Dataset Provided to HRW


Percent of Cases missing from DHS Publications














































Nationalities deported

Individuals representing 184 different nationalities were deported after serving their sentences for criminal convictions between April 1, 1997, and August 1, 2007. Mexicans were by far the largest national group, representing 78.2 percent or 701,700 of those deported, which is not surprising since individuals of Mexican national origin represent the largest percentage-27.9 percent (or 11.5 million)-of the 37.5 million persons in the foreign-born population in the United States.[38] The vast majority of deportees on criminal grounds (97.1 percent of the total) can be grouped into 25 nationalities, each of which had more than 1,000 deportations, or an average of over 100 deportations per year. Online Appendix F presents the full list of nationalities deported.[39]

Table 3: Nationalities with Greater than 1,000 Deportees




Cumulative Percent












El Salvador





Dominican Republic

































































Trinidad and Tobago






























South Korea














Countries receiving deportees

There were 1,845 non-citizens in the dataset who were deported from the United States on criminal grounds for whom ICE failed to record data on the country to which they were sent. The dataset also includes 149 individuals deported to the USSR, a country that by 1997 (the earliest date of deportations included in the data) had not existed for over five years.[40] There are also eight deportees in the dataset labeled as having been deported to the US, indicating a clear data management error.

In the overwhelming majority of cases, the deportee was returned to his or her country of nationality. However, a total of 3,194 deportees were deported to a country other than that listed as their country of nationality. This may represent individuals with dual citizenship, but the ICE dataset only provided one nationality for each person deported, perhaps leaving off the individual's second country of nationality. It may also point to a pattern of deporting people to places to which they hold no citizenship ties. Individuals representing 133 different nationalities were deported to a country other than their country of nationality. Countries with the greatest numbers of citizens deported to other countries or territories were Canada (992), the UK (400), Mexico (220), the Netherlands (186), and France (124).[41]

Data on Immigration Status

Data deficiencies

Although there have been occasional cases in which ICE erroneously deported US citizens from the United States, the vast majority of individuals deported for criminal conduct are in fact non-citizens.[42] However, non-citizens in the United States are assigned one of a wide variety of immigration statuses, which determines the legality of their presence in the country and under what terms and conditions they may remain. Thus, the immigration status of each person deported must be known in order to fully understand what rights and interests he or she may have.

It is of particular concern that there are 62,480 individuals, representing 7 percent of the total, in the ICE dataset with an "unknown" immigration status (which includes individuals coded as "other," "unknown or not reported," or "withdrawal," as well as those with no immigration status code). This is of concern because immigration status determines what an individual's rights are in the deportation process. We have no way of knowing whether the agency simply failed to document their status and enter it into its data management system, or whether the agency truly was unable to place each of these non-citizens in an immigration status category.

In the deportation process, all procedures, penalties, and possible defenses to deportation stem from an individual's immigration status. This means that with respect to 62,480 persons, there are serious concerns as to whether human rights, immigration, and/or constitutional law violations occurred in these individuals' deportation cases. It is possible that ICE recorded many non-citizens' immigration status as "other" because it lacked sufficient documentation to confirm their immigration status. The fact that 18 percent, or 11,246 of the deportees in the "unknown" category (which includes persons coded as "other") had been convicted of false citizenship may corroborate this hypothesis for at least a segment of the total. However, we note that ICE was able to record nationalities for all deportees in the "unknown" immigration status category.

Most common immigration statuses among deportees

In total, the people deported on criminal grounds in the dataset held 84 different immigration statuses. However, the vast majority (98.4 percent of the total) held one of six statuses (see Table 4; see Online Appendix G for a frequency table for all immigration statuses).[43] A majority of deportees, 73 percent or 655,145, held the immigration status of "without inspection," meaning that they entered the country without being inspected by an immigration official at a border crossing or another port of entry.

Table 4: Six Most Common Immigration Statuses

Immigration Status




Without Inspection

These individuals entered the United States without being inspected by a US border official at a border crossing or another port of entry (that is, an airport or a seaport).




These individuals are lawful permanent residents, or green card holders. The lawful permanent resident status allows for an unlimited lawful presence in the United States, allows these individuals to work legally, and allows for eventual citizenship through naturalization.




Coded as "other," "withdrawal," or "unknown or not reported" by ICE.




These individuals have been granted time-limited, but renewable, permission to remain in the United States. Parolee status is granted under the discretion of the Attorney General and often, though not always, is accompanied by legal permission to work in the United States. The Cubans who entered the United States through the Mariel boatlift in 1980 are an example of parolees.



Visitor for Pleasure

These are non-citizens legally inside the United States until the expiration of their time-limited tourist visas.



Expedited Removal Alien

These are non-citizens who have been subject to an expedited process because they have been apprehended within 100 miles of the border, or have arrived at a port of entry without valid entry documents and have made a request for asylum from persecution or protection against return to torture. They have permission to remain in the United States until their claims have been heard in the expedited removal hearings process.






Legal versus illegal immigration status

We have placed all 84 immigration statuses into one of six categories based on the legality of the non-citizen's presence in the United States. A table provided at Online Appendix H provides an explanation of legality for each of these 84 immigration statuses.[44] The 179,038 individuals (constituting 20 percent of the total number of non-citizens in the dataset) who were legally present in the US and were subsequently deported on criminal grounds after serving their criminal sentences are of particular importance from a human rights perspective, as this group (as emphasized in Chapter IV) has the strongest rights claims against summary deportation.


This group is also worth close examination because such an examination counters alarmist and ill-informed statements giving the impression that deportation policies focus exclusively on people who are illegally in the country and who commit violent crimes. An example of such a claim was made by Representative Steve King (R-Iowa), citing statistics without sources:

[I]f we would have enforced our domestic laws so when people violated immigration laws internally, domestically; if we did those things, then we wouldn't have illegal aliens in America to commit the crimes. And that would equate and extrapolate down to 12 fewer murders every day, 13 fewer people that die at the hands of negligent homicide, primarily the victims of drunk drivers, at least 8 little girls that are victims of sex crimes on a daily basis, and that number could be well higher than that … This is a slow-rolling, slow-motion terrorist attack on the United States costing us billions of dollars and, in fact, thousands of lives, and we have an obligation to protect the American people, and that means seal and protect our borders.[45]

Not only do the deportation laws sweep up people legally and illegally present alike, most of those deported have not committed violent offenses, as will be demonstrated below.[46] Similarly, Bill O'Reilly often lumps criminality and illegal presence together in his Fox News television show, The O'Reilly Factor. In May 2007, with reference to several incidents of criminal investigations of non-citizens, he said, "The problem of criminal illegal aliens is now at a tipping point in the USA," and that there is "anarchy" in the immigration zone, with the government "doing little."[47] In fact, the government appears to be doing more than targeting "criminal illegal aliens." The totals presented in Table 5 below show that at least one-fifth of those deported under these laws were in the US legally.

Table 5: Legality of Immigration Status




Illegally Present



Legally Present









We have further analyzed the "legally present" category to highlight the differences within this category between individuals with time limits on their stays within the US and those with adjustable or renewable statuses. As Table 6 reveals, nearly half, or 89,426 of those with a "legally present" immigration status had no time limits on their stay. Thirty-one percent of those with a "legally present" immigration status had a finite time limit on their stay. It is unknown how many of these individuals had overstayed their visa at the time of arrest or deportation. The other 18.9 percent of those in the "legally present" immigration status category had time limits but had either a renewable or adjustable status that would have enabled them to change their status if the Bureau of Citizenship and Immigration found in their favor.

Table 6: Legality of Immigration Status

Immigration Status



Illegally Present



Legally Present, no time limit on stay, renewable/adjustable status



Legally Present, with time limit on stay



Legally Present, with time limit on stay, but renewable/adjustable status



Legally Present, with time limit on stay, but can adjust status if court finds in favor









Of those with a "legally present" immigration status, 49 percent held the immigration status of "immigrant," which means that the individual was a lawful permanent resident or a "green card" holder, a status with no time limit on stay in the United States. Table 7 displays the eight most common immigration statuses in the "legally present" category, comprising 96.5 percent of all legally present non-citizens.

Table 7: Individuals in "Legally Present" Immigration Status Category (n>1,000)

Immigration Status


Percent of those in "Legally Present" category

Cumulative Percent

Legality of Status





Legally present, no time limit on stay, renewable / adjustable status





Legally present, with time limit on stay, but renewable / adjustable status

Visitor for Pleasure




Legally present, with time limit on stay

Expedited Removal Alien




Legally present, with time limit on stay

Visitor for Business




Legally present, with time limit on stay





Legally present, with time limit on stay

Visitor without Visa 90 days




Legally present, with time limit on stay





Legally present, no time limit on stay, renewable / adjustable status

Data on Criminal Conduct forming Basis for Deportations

Data deficiencies

The ICE dataset contained up to five criminal conviction codes for each individual non-citizen. In total, non-citizens in the dataset were arrested for 356 distinct crimes (see Appendix A for a frequency table of criminal offense codes). We categorized all of these crimes into one of six categories, by cross-referencing with the National Crime Information Center codebook:

  • Offenses involving violence against persons,
  • General offenses with potential to cause harm,
  • Non-violent drug offenses,
  • Non-violent general offenses,
  • Non-violent immigration offenses, and
  • Non-violent theft offenses.

These categories were ranked for level of seriousness. Out of the entire dataset, 24 percent (or 215,308 cases) had data for more than one crime committed. When individuals in the dataset were convicted of more than one crime, we used the deportee's most serious crime for our analysis.

Surprisingly, 395,272 (44 percent) of the cases contain no crime data. We are concerned about this result for much the same reason we are concerned to find 7 percent of cases containing no immigration status information. Obviously, each non-citizen's criminal conduct is an extremely important factor in determining his or her rights and defenses to deportation under US immigration law. Moreover, we are particularly disturbed because the dataset provided to us was specifically produced by ICE in response to our request for "individual level case-by-case records for each non-citizen removed on criminal grounds" (see Appendix B for our amended request letter to ICE). We defined "non-citizen removed on criminal grounds" through reference to the 40 sections of the Immigration and Nationality Act that enumerate all types of criminal conduct that can render someone subject to deportation from the United States. The failings evident in the ICE data management system-particularly when data were produced in response to a request for records specifically about deportees with criminal conduct, and nevertheless no information about criminal conduct was recorded for 44 percent of cases-should be of serious concern to both the Department of Justice and Congress, and inquiry should be made as to whether individuals have been deported without regard to criminal record.


Our concerns about the lack of crime data prompted us to write to ICE on October 3, 2008, presenting these data deficiencies and offering the agency an opportunity to provide us with any explanations or clarifications (see Appendix D). The agency responded to our concerns in a letter dated February 2, 2009, stating, "we can report that ICE is in the process of improving its data management systems to more consistently record criminal conviction codes (NCIC codes) for all aliens removed from the United States with criminal convictions." The letter goes on to explain that the previous data management system was retired in August 2008 and replaced with a new system that ICE officers have attended trainings on, a "Data Quality and Integrity Unit" has been set up, and internal policy guidance on data entry, including on criminal history, was distributed to ICE staff through December 2008.[48]

Background on criminal conduct forming basis for deportations

Under US immigration law there are two broad categories of criminal conduct that can form the basis for an individual's deportation: aggravated felonies and crimes of moral turpitude. Many types of crime fit under these two broad headings. Since immigration law was changed in 1996, aggravated felonies include the following broad categories of crime:

  • any crime of violence (including crimes involving a substantial risk of the use of physical force) for which the term of imprisonment is at least one year,
  • any crime of theft (including the receipt of stolen property) or burglary for which the term of imprisonment is at least one year, and
  • illegal trafficking in drugs, firearms, or destructive devices.[49]

The following specific crimes are also listed as aggravated felonies:

  • murder,
  • rape,
  • sexual abuse of a minor,
  • illicit trafficking in a controlled substance, including a federal drug trafficking offense,
  • illicit trafficking in a firearm, explosive, or destructive device,
  • federal money laundering or engaging in monetary transactions in property derived from specific unlawful activity, if the amount of the funds exceeded $10,000,
  • any of various federal firearms or explosives offenses,
  • any of various federal offenses relating to a demand for, or receipt of, ransom,
  • any of various federal offenses relating to child pornography,
  • a federal racketeering offense,
  • a federal gambling offense (including the transmission of wagering information in commerce if the offense is a second or subsequent offense) that is punishable by imprisonment of at least one year,
  • a federal offense relating to prostitution,
  • a federal offense relating to peonage, slavery, involuntary servitude, or trafficking in persons,
  • any of various offenses relating to espionage, including protecting undercover agents or classified information, sabotage, or treason,
  • fraud, deceit, or federal tax evasion, if the offense involves more than $10,000,
  • alien smuggling, other than a first offense involving the alien's spouse, child, or parent,
  • illegal entry or reentry of an alien previously deported on account of committing an aggravated felony,
  • an offense relating to falsely making, forging, counterfeiting, mutilating, or altering a passport or immigration document if (1) the term of imprisonment is at least a year and (2) the offense is not a first offense relating to the alien's spouse, parent, or child,
  • failure to appear for service of a sentence, if the underlying offense is punishable by imprisonment of at least five years,
  • an offense relating to commercial bribery, counterfeiting, forgery, or trafficking in vehicles with altered identification numbers, for which the term of imprisonment is at least one year,
  • an offense relating to obstruction of justice, perjury or subornation of perjury, or bribery of a witness, for which the term of imprisonment is at least one year,
  • an offense relating to a failure to appear before a court pursuant to a court order to answer to or dispose of a charge of a felony for which a sentence of two years' imprisonment or more may be imposed, and
  • an attempt or conspiracy to commit one of the foregoing offenses.[50]

While some of these aggravated felonies would seem to be severe offenses for which deportation is an appropriate punishment, in practice it is not always clear cut. For example, Ramon H.[51] (a pseudonym) is originally from Mexico. He married a United States citizen, Pamela H. (a pseudonym), in 1990. In February 1993 Ramon pled guilty to lewd or lascivious acts with a minor. After his plea, he completed his sentence of probation, according to his probation officer, "in an exemplary fashion."[52] Ramon H. applied to adjust his status to that of a lawful permanent resident through his US citizen wife in 1996, but in 2001 the Department of Homeland Security informed him that he was deportable for his criminal conviction, and he was placed in removal proceedings in August 2004.

The circumstances of Ramon H.'s crime were later described by his niece Kelda in a sworn affidavit that she submitted during his deportation hearing. Kelda explained that during a family gathering, her uncle Ramon patted her "lightly on the butt … for no apparent reason."[53] Kelda mentioned the incident to a friend at school, who in turn told a teacher, and the school called the police, resulting in Ramon H.'s conviction and order of deportation.

Ricardo S. also faced separation from his US citizen wife and two children because of an aggravated felony drug conviction.[54] He was ordered deported because of a conviction for possession with intent to distribute of a small amount of heroin, for which he was advised by a defense attorney to plead guilty, and in return he received no jail time but was ordered to pay a fine of $500 and serve two years probation, which he completed without incident. Ricardo S. had no other criminal convictions and worked in construction in the Chicago area. His conviction was brought to the attention of the immigration authorities because he and his US citizen wife, who were married in 2001, applied to adjust Ricardo S.'s status to that of a lawful permanent resident. Looking back on his one conviction, Ricardo S. said,

I feel bad about it because of my family. If I was by myself, without my wife or any children, it would have been a lot different. But I feel real bad for them…. Maybe if they would have caught me with a ton of drugs [I could understand them wanting to deport me], or if I ever murdered somebody. But it was the only one…. I wish that [when he applied for his green card] they would have just told me I didn't qualify. I have kids who are citizens and a wife who is a citizen but I wish they would have just let me continue working to support my family….[55]

Non-citizens are also deportable if they are convicted of a "crime involving moral turpitude" within five or in some cases 10 years after they enter the United States and their crime carries a sentence of one year or longer.[56] A non-citizen is also deportable if she is convicted of two or more crimes of moral turpitude at any time after admission.[57] In 1997 Congress did not change the crimes considered to meet the definition of "moral turpitude." However, it did make it more difficult for non-citizens with convictions for crimes of moral turpitude to defend against deportation.

For example, Mark Ferguson, a native of the United Kingdom who had lived in the United States lawfully as a green card holder since the age of three, was convicted of two or more crimes of moral turpitude for "mooning" (showing his nude buttocks to) women.[58] Ferguson testified that in the past he had mooned a woman about once every six months, but was under psychiatric treatment for the practice, and under treatment had not reoffended for two years. He submitted expert testimony that he was not sexually aroused by the practice, had an "unusually low" chance of reoffending, and had strong family connections to the United States, including because he was a primary caregiver for his deceased sister's children. The Board of Immigration Appeals found that although he was statutorily eligible for waiver ("cancellation of removal") under INA Section 240A, cancellation of Ferguson's removal would not be in the interests of the United States. On appeal, the court found that it had no power to review that discretionary decision.[59]

Types of crime forming basis for deportations

Human Rights Watch analyzed the 356 crime codes provided to us by ICE and classified each into one of six categories. Appendices C and E provide more details on the crimes that fit within each of these categories. Table 8 and Figure 2 provide information on the frequency with which individuals were deported from the US for crimes falling into each of these six categories.

Table 8: Total Cases, including Cases with No Crime Data

Crime Category Forming Basis for Deportation



No Crime Data



Non-Violent Immigration Offense



Non-Violent Drug Offense



Non-Violent General Offense with Potential to Cause Harm



Offense involving Violence Against Persons



Non-Violent Theft Offense



Non-Violent General Offense






Figure 2: Crime Category Forming Basis for Deportation (excluding cases with no crime data)

If we combine "general offenses with potential to cause harm" with "offenses involving violence against persons" to create a "violent offenses" category, we see in Table 9 and Figure 3 that the vast majority of deportees for whom we have crime data (72.2 percent) were deported for non-violent crimes; only 27.8 percent were deported for violent or potentially violent offenses.

Table 9: Violent v. Non-Violent Offenses

Offense Type



Non-Violent Offense



Violent or Potentially Violent Offense






Figure 3: Violent v. Non-violent Offenses

Crime Data Combined with Immigration Status

Data deficiencies

Examining all 897,099 cases by the legality of the individual's immigration status and the offense for which he or she was deported can give a general sense of the types of non-citizens being deported from the United States for different kinds of crimes. It is significant that only 5.8 percent of legally present non-citizens in the data set were listed as deported for a violent or potentially violent offense. However, it is of serious concern that 74.8 percent of those listed in the data set as legally present were deported without any crime data recorded. This raises the question as to whether there is a serious problem in data recording practices, or deportation practices, or both. Without accurate crime data, we must also raise the possibility that some of these people were unlawfully deported in violation of both US and international human rights law. Moreover, without accurate crime data, the US public and government cannot know exactly how many legal, long-term residents or other legally present non-citizens have been deported from the United States for crimes that are petty or serious.

Non-citizens who are legally present are the group most likely to have serious human rights claims against summary deportation, and while they represent 20 percent of all those deported, they (disproportionately) represent 33.9 percent of all those deported without crime data. ICE failed to record any crime data for 94.9 percent of lawful permanent residents (actual green card holders). Our concern with the failure to record crime data is not a mere question of poor recordkeeping: It is based on the fact that for those non-citizens who were legally in the country, certain criminal convictions would form the only legal basis for their deportations, raising the question as to whether these people were deported unlawfully.

Thirty-seven percent of illegally present individuals recorded in the dataset as deported also do not have any crime data. (Although this too raises serious questions about the data management capacities of ICE, one possible explanation for at least some of these omissions is that these individuals were deported solely on the basis of their undocumented status, but without any allegations or evidence of criminal conduct. However, if this were the case, it still poses the question why these hundreds of thousands of persons were included in a dataset specifically produced to contain only data relevant to persons deported on criminal grounds.) While illegally present individuals account for 73 percent of all those deported, they (disproportionately) account for 89 percent of all individuals deported for a violent or potentially violent offense. It is possible that these percentages may be skewed because of the large number of cases in all immigration status categories without crime data. Nonetheless, using the ICE data, it appears that illegally present individuals are deported for violent or potentially violent offenses at a greater rate than legally present individuals (see below, section "The Seriousness of Criminality within All Immigration Status Categories").

There is clearly a difference between how well crime data are recorded for individuals with different immigration statuses. The disparity is at its greatest when we examine those who are illegally present-those with a "without inspection" or "stowaway" immigration status code-versus those who are legally present. Those holding an "illegally present" immigration status code are actually one of the groups with the greatest probability of having their criminal conduct documented. In the entire dataset, 56 percent of deportees had crime data. For illegally present non-citizens, this number increases to 62.6 percent.

The trend reverses for those in the "legally present" category. Only 25.2 percent of cases with a "legally present" immigration status have crime data.

We chose to do further analysis on the data corresponding to the three legally present immigration statuses of "immigrant," "parolee," and "refugee." This is because the deportations of persons in these categories raise the greatest human rights concerns: "immigrant" (because they are lawful permanent residents), "parolee" (because they are legally in the country, in most cases for humanitarian reasons), and "refugee" (because they are legally in the country due to fears of persecution at home). Unfortunately, we have discovered that these three immigration statuses of most concern have even less accurate crime data. As shown in Table 10 below, only 5.1 percent of those with an "immigrant" status, 9.4 percent of "refugees," and 27.5 percent of "parolees" have crime data.

Table 10: ICE Management of Crime Data

Cases Included

Percent of cases with crime data

Percent of cases without crime data

Frequency of cases with no crime data

"Immigrant" status code




"Parolee" status code




"Refugee" status code




Immigrant, Parolee, and Refugee combined




All Immigration statuses in "Legally Present" category




"Without Inspection" Immigration Status Code




"Unknown" status code




All Cases




If, as was requested by Human Rights Watch, the dataset recorded everyone deported for some sort of criminal conduct, ICE systematically failed to record crime data for those who were legally in the country (especially those in the "immigrant," "parolee," and "refugee" categories). The extent of the difference in how often crime data were recorded between this group and the illegally present group implies that there is some sort of institutional dysfunction at work. In fact, these three types of legally present deportees combined were 14 times less likely to have crime data recorded than illegally present deportees.[60]

The immigration status category of "unknown" is also particularly problematic. As noted above, individuals in this category held one of three immigration statuses, "other," "unknown or not reported," or "withdrawal." Of these individuals, 26.2 percent have no crime data, meaning there were 16,364 people deported for a criminal offense for whom we not only do not know their immigration status, but ICE also made no record of the crime for which they were deported. This "double unknown" of immigration status and criminal offense highlights extraordinary gaps in ICE data management.

The Seriousness of Criminality within All Immigration Status Categories

By removing all cases without crime data in the ICE dataset, we see that the vast majority of deportees were deported for a non-violent offense. In total, across all immigration status categories, more than two-thirds of those for whom we have crime data were deported for a non-violent crime-70.5 percent were deported for a non-violent offense and 29.5 percent were deported for a violent or potentially violent offense.

As Figure 4 shows, illegally present non-citizens were more likely to have been deported for a violent or potentially violent offense and less likely to have been deported for a non-violent offense than legally present non-citizens. Since the laws allowing for these deportations were passed with a clear focus on those responsible for violent offenses, the higher percentages of deportation for violent offenses among undocumented persons raises the important policy question why enforcement resources are not focused exclusively upon persons present in the United States in an undocumented or illegal status, who were also involved in serious, violent criminal offenses.

Figure 4: Criminality by Immigration status

Immigration status: Legally present

Although the "legally present" immigration status category represents 20 percent of all cases, it (disproportionately) represents 33.9 percent of all cases with no crime data. As noted above, of those in the "legally present" category, 74.8 percent have no crime data.

It is both confusing and problematic that the crime code most frequently recorded for the "legally present" category was "illegal entry" (see Appendix E). This implies that even though these non-citizens were in the country with a legal immigration status, they were convicted of the crime of "illegal entry." We can only speculate as to why this would be the case. It may be that a person who originally was allowed to enter the country in a legal status was subsequently discovered to have falsified information that retroactively made his or her entry illegal, or that his or her legal status did not permit multiple trips out of the US, making any subsequent entry "illegal," although it seems unlikely that either of these scenarios would have occurred in more than 10,000 cases. Alternatively, it may be the case that these anomalies are due, once again, to data management failures by ICE.

Nevertheless, for deportees in the "legally present" category for whom we do have crime data, it is significant that 77 percent of them were deported for non-violent offenses, as shown in Figure 4. A more detailed description of those offenses is provided in Appendix E.

Immigration status: Lawful permanent resident

Representing 49 percent of those in the "legally present" category, 87,844 individuals hold the immigration status of "immigrant," which is the status for persons in the United States as lawful permanent residents (green card holders). As noted above, the vast majority, 94.9 percent, of those with an "immigrant" status have no crime data: Although these non-citizens have been afforded the most "privileged" immigration status available in the United States by immigration authorities, no one seems to have recorded the underlying criminal basis for their deportation.

Although only 5 percent of individuals holding an "immigrant" status and recorded in the dataset as deported have crime data, this still allows us to examine the criminal convictions of 4,453 non-citizens in this status grouping, which is valuable to analyze because of the sheer numbers involved. Table 11 shows that of this subgroup of "immigrant" status with crime data, the large majority, or 68 percent, were deported from the United States for non-violent offenses. Appendix E gives more detailed information on the criminal conduct forming the basis for deportations of persons in the "immigrant" status category.

Table 11: Immigration Status "Immigrant" -Type of Crime (excluding cases with no crime data)



Non-Violent Offense



Violent or Potentially Violent Offense






Immigration status: Parolee

The immigration status of "parolee" is used for individuals who have been granted time-limited but renewable permission to remain in the United States. "Parolee" status is granted at the discretion of the Attorney General and often, though not always, is granted to persons with humanitarian reasons for not being able to return to their home countries. In addition, it is often, though not always, accompanied by legal permission to work in the United States. The ICE dataset contained no crime data for 21,398, or 72.5 percent, of all parolees. For those who do have crime data, in 79.5 percent of cases a non-violent crime formed the basis for their deportation from the United States, as shown in Table 12. Appendix E gives more detail on the most common crimes forming the basis for deportations of parolees.

Table 12: Immigration Status "Parolee" -Type of Crime (excluding cases with no crime data)



Non-Violent Offense



Violent or Potentially Violent Offense






Immigration status: Refugee

There were 1,038 deportees with the immigration status of "refugee." Refugees may apply for legal permanent residence in the United States after one year of residence. Refugees, like immigrants and parolees, are individuals with serious human rights interests at stake when they are facing deportation on criminal grounds (US obligations in this regard, as a party to the 1967 Protocol Relating to the Status of Refugees, are discussed in Chapter IV). Again, the vast majority of refugees deported, 90.6 percent, did not have crime data recorded by ICE.

As is perhaps obvious, the deportations of refugees on criminal grounds raise serious human rights concerns because their removals from the United States raise questions of life and death. Unfortunately, due to the restrictive laws put in place in the United States in 1997, refugees facing deportation on criminal grounds are often barred from raising their fears of persecution during their deportation hearings. The crimes for which some have been deported are not serious enough to deprive the person of refugee status under the Refugee Convention, and therefore in accordance with refugee law they should be protected from return. Nevertheless, because US law does not allow them to raise these fears in a deportation hearing, an unknown number of them in fact may have been returned to places where they were subjected to persecution. US law does allow for persons to raise concerns that they will face a real risk of torture prior to deportation, regardless of their criminal conviction.

However, there are many refugees who fear persecution but not torture-for example, an outspoken member of the political opposition might fear being imprisoned without trial if he were deported, which is a form of persecution but not torture. The United States is regularly violating these refugees' rights by deporting them for criminal convictions without first providing a fair hearing on their fears of persecution, and protecting them from return if those fears are proved valid. Table 13 shows that 62 percent of refugees for whom we have crime data were deported under such perilous conditions for a non-violent offense, and 38 percent were deported for a violent or potentially violent offense. Appendix E gives more details on the most common offenses forming the basis for the deportations of refugees from the United States.

Table 13: Immigration Status "Refugee" -Type of Crime (excluding cases with no crime data)



Non-Violent Offense



Violent or Potentially Violent Offense






Immigration status: Expedited removal pending credible fear

The immigration status of "expedited removal pending credible fear" raises concerns similar to those presented by refugees. Persons holding this immigration status can be considered applicants for refugee status, since they were placed in summary removal procedures, but pending a credible fear interview. Credible fear interviews are the first step that persons who flee to the United States because of a fear of persecution must undergo.

While the persons in this category did not have their status resolved prior to the time of deportation from the United States, it can be assumed that they all raised fears of persecution with immigration authorities, and that for some percentage those fears were well-founded, making them genuine refugees. Table 14 shows that 76.7 percent of these people for whom we have crime data were deported for non-violent offenses. Appendix E gives more detail on the most common offenses forming the basis for the deportations of persons awaiting their credible fear interviews.

Table 14: Immigration Status "Expedited Removal Pending Credible Fear" -Type of Crime (excluding cases with no crime data)



Non-Violent Offense



Violent or Potentially Violent Offense






Immigration status: Illegally present

For this report, the 436 deportees with the immigration status of "stowaway" were combined with those with the "without inspection" immigration status to make up the "illegally present" category.[61]

Of all cases in the ICE dataset, 73 percent (or 655,145) of the deported non-citizens held the immigration status of "without inspection," meaning that they entered the country in an undocumented status without being inspected by an immigration official at a port of entry. Of these individuals recorded in the dataset as deported, 244,804, or 37.4 percent, do not have crime data. Of all cases without crime data, 61.9 percent have the immigration status of "without inspection." In contrast to the other instances described above, in which ICE's failure to include crime data is of serious concern, the failure to include crime data for those who entered without inspection may have a plausible explanation (as discussed above, subsection "Types of crime forming basis for deportations"): these non-citizens could have been deported simply on the basis of their undocumented status alone.

In fact, of those non-citizens with an illegally present immigration status who do have crime data, 24.1 percent, or 98,940, were convicted of the crime of "illegal entry." In other words, not only were these persons ordered deported because they entered the United States in an undocumented status, which is enough under US law to deport them, but in addition, they were convicted of the federal crime of "illegal entry" and sentenced to criminal punishment prior to their removal. Tables 15 and 16 show the types of crimes and general categories of offenses for persons illegally in the country and subsequently deported after criminal conduct.

Table 15: Offenses in Illegally Present Immigration Status Category

Violent vs. Non-Violent Offense



Non-Violent Offense



No Crime Data



Violent or Potentially Violent Offense






Table 16: Crime Categories, Illegally Present Immigration Status (excluding cases with no crime data)

Crime Category



Non-Violent Immigration Offense



Non-Violent Drug Offense



Non-Violent General Offense with Potential to Cause Harm



Offense Involving Violence Against Persons



Non-Violent Theft Offense



Non-Violent General Offense






For those people with crime data and an illegally present immigration status, 69.7 percent were deported for a non-violent offense. Individuals with a "without inspection" immigration status were most often deported for a non-violent immigration crime. In fact, 98,940 individuals holding this status were deported for the non-violent immigration crime of "illegal entry." Appendix E shows the 10 most common offenses forming the basis for the deportation of those with an illegally present immigration status.

Of those with an unknown immigration status, 49.7 percent were convicted of one of three immigration offenses. This may imply that they were in the United States illegally, despite ICE's failure to record an immigration status for them. Appendix E provides more detail on the most common criminal offense codes for people in the unknown immigration status category.

[34] Aggregate numbers of non-citizens deported on criminal grounds have been published by DHS on its website for many years. The figures for 1997 through 2006 (the most recent year for which data are available through these sources), show that a total of 768,345 non-citizens were removed on criminal grounds during that period.

[35]There has been much speculation about an increase in the use of immigration enforcement powers by ICE officials since the September 11 terrorist attacks on the United States. While the dataset does reveal an increase in the number of non-citizens deported on criminal grounds since September 11, 2001, it is difficult to know how much of this is attributable to increased enforcement motivated by the terrorist attacks, and how much to other initiatives by ICE to increase enforcement more generally. In the 53 months between April 1, 1997, and September 10, 2001, 340,882 people were deported on criminal grounds-a monthly average of 6,431. In the 71 months between September 11, 2001 and August 1, 2007, 556,217 people were deported-a monthly average of 7,834. Before September 11 an average of 210 people were deported per day; after September 11 an average of 259 were deported per day. Of the 100 days with the highest number of deportations, 99 occurred after September 11, 2001.

[36]See, for example, U.S. Department of Homeland Security, Office of Immigration Statistics, "2007 Yearbook of Immigration Statistics," http://www.dhs.gov/xlibrary/assets/statistics/yearbook/2007/ois_2007_yearbook.pdf (accessed March 20, 2009).

[37] Due to rounding, for Table 2 and all subsequent tables, numbers may not add up to 100 percent.

[38]Pew Hispanic Center, "Tabulations of the 2006 US Census American Community Survey," January 2008, Table 7, http://pewhispanic.org/files/factsheets/foreignborn2006/Table-7.pdf (accessed March 26, 2009).


[40]It is possible that the ICE data here reflected only available passport data and not actual country of return; USSR passports remained valid for years beyond the demise of the Soviet Union, pending the transfer to national passports in ex-Soviet countries.

[41]Human Rights Watch wrote to the governments of the United States, Canada, France, Mexico, the Netherlands, and the United Kingdom between January 13 and March 10, 2009, raising our concerns about these removals and asking for additional information. On March 18, 2009 we received a response from the government of Mexico, indicating that some of the information Human Rights Watch shared in our letter "undoubtedly has serious human rights implications" ("sin duda alguna tiene serias implicaciones en el ámbito de los derechos humanos") but that the Mexican government has "no knowledge of any case similar to those mentioned" [by Human Rights Watch, that is, cases in which Mexican citizens were deported to a third country] ("Hasta el momento no se tiene conocimiento de algún caso de las dimensiones como el que usted menciona."). Letter from Juan Manuel Gómez Robledo, undersecretary for multilateral affairs and human rights, Department of Foreign Relations, Government of Mexico, to Human Rights Watch, March 18, 2009. As of this writing, we have not received a response from the United States, Canada, France, the Netherlands, or the United Kingdom.

[42] "Deported U.S. Citizen is Returned to Family," Associated Press, August 8, 2007.

[43] http://www.hrw.org/sites/default/files/reports/appendixg.pdf.


[45]Representative Steve King (R-Iowa), "Comparing the Statistics," Statement to the U.S. House of Representatives, May 3, 2006, http://www.house.gov/apps/list/speech/ia05_king/sp_20060503_stats.html (accessed March 20, 2009).

[46] See discussion in subsection "Types of Crime forming Basis for Deportations," below.

[47] Bill O'Reilly, "The Problem of Criminal Illegal Aliens is Now at the Tipping Point," The O'Reilly Factor, Fox News, May 9, 2007, http://www.foxnews.com/story/0,2933,270974,00.html (accessed March 20, 2009).

[48] Letter from James T. Hayes, Jr., director, Office of Detention and Removal Operations, U.S. Immigration and Customs Enforcement, to Human Rights Watch, February 2, 2009.

[49] Immigration and Nationality Act, Section 101(a)(43), subsections (B)(C)(F)(G); 8 U.S.C. Section 1101(a)(43), subsections (B)(C)(F)(G).

[50]Immigration and Nationality Act, Section 101(a)(43), subsections (A)-(U); 8 U.S.C. Section 1101(a)(43), subsections (A)-(U).

[51]This illustrative case example also appeared in our 2007 report, Forced Apart , pp. 20-21, 61-62.

[52] Letter from Dick Tschinkel, Los Angeles County Probation Department, October 15, 2004, on file with Human Rights Watch.

[53] Affidavit of Kelda O. (pseudonym), submitted in opposition to Ramon's deportation, October 15, 2004, on file with Human Rights Watch.

[54] Human Rights Watch interview with Ricardo S. (pseudonym), Chicago, Illinois, February 3, 2006. This illustrative case example also appeared in our 2007 report, Forced Apart, pp. 21-22.


[56] 8 U.S.C. Section 1227(a)(2)(A)(i)(I) and (II).

[57]8 U.S.C. Section 1227(a)(2)(A)(ii).

[58]This illustrative case example also appeared in our 2007 report, Forced Apart, p. 23.

[59]Ferguson v. Attorney General of the United States, 2007 U.S. App. LEXIS 3100 (3d Circuit, February 9, 2007).

[60]Deportees holding the three immigration statuses of "refugee," "immigrant," or "parolee" were placed in one "legally present combined" category, and illegally present deportees were placed in a second category. We ran a logistic regression odds ratio test to see whether there is a significant difference in the recording of crime data between the two categories. Using the presence of crime data as the output variable, the test leaves no doubt that there is a statistically significant difference between the two categories' likelihood of having their crime data recorded by ICE (chi squared = 119008.69, z=-270.81 (p > .000). The odds ratio provided by this test proves with 95 percent certainty that illegally present deportees were between 13.7 and 14.2 times more likely to have crime data recorded than those in the "legally present combined" category.

[61] We recognize that some additional persons with time-limited legal statuses, currently grouped in the legally present category, may have overstayed their visas, thereby transforming their status from legally present to illegally present. Nevertheless, the ICE dataset recorded these persons as continuing to hold a legal, albeit time-limited status. In addition, if they had simply overstayed their visas, they could have been deported for that reason alone and there would be no reason to include them in a dataset of individuals deported on criminal grounds. For these reasons, we have grouped all such persons in the legally present category.