Very timely case-by-case data from Immigration and Customs Enforcement (ICE) show that no more than 14 percent of the "detainers" issued by the government in FY 2012 and the first four months of FY 2013 met the agency's stated goal of targeting individuals who pose a serious threat to public safety or national security. In fact, roughly half of the 347,691 individuals subject to an ICE detainer (47.7 percent) had no record of a criminal conviction, not even a minor traffic violation.
The very recent ICE data further showed that if traffic violations (including driving while intoxicated) and marijuana possession are put aside, fully two thirds of all detainers had no record of a conviction. See Figure 1.
This thoroughly-documented government enforcement effort sharply contrasts with the multiple press releases and official statements issued by the agency. "The removal of aliens who pose a danger to national security or a risk to public safety shall be ICE's highest immigration enforcement priority," claimed former ICE Director John Morton in an agency-wide memo describing the Obama Administration's approach to immigration enforcement announced in June 2010. The stark difference between the agency's rhetoric and its actual performance is startling.
Detainers, often called "immigration holds," are a primary tool that ICE uses to apprehend the suspects it is seeking. The official ICE notices ask local, state and federal law enforcement agencies not to release suspected non-citizens held at their facilities in order to give ICE an opportunity to take them into its custody. However, there is no legal obligation for state and local law enforcement agencies to honor these detainers, and some jurisdictions have begun refusing to hold arrestees for minor violations or those who pose little risk to public safety.
The extensive new information presented in this report about ICE detainers and how the agency has used them was obtained under the Freedom of Information Act (FOIA) and analyzed by the Transactional Records Access Clearinghouse (TRAC) at Syracuse University. These data cover all ICE detainers placed during the sixteen month period from October 1, 2011 to January 31, 2013.
(This is the second in a three-part series discussing ICE detainers. Part I of the series, published July 25, covered monthly trends in detainer use and found that the average monthly number of detainers issued this fiscal year was down 19 percent from the monthly average level during FY 2012. It also examined which ICE programs had issued them. Part III of the series, now in preparation, will compare data on a state-by-state and a detention facility-by-detention facility basis to examine where ICE detainers were sent. TRAC's February report on detainers covered earlier time periods and was based on much more limited information.)
Who Is Being Targeted: Reasons Cited When Detainer Placed
The boxes checked on the detainer forms indicate the reason(s) DHS officers are asking the law enforcement agencies to hold the arrestees. In its FOIA request to ICE, TRAC asked for the information recorded in these boxes both on the old detainer form in use until December 21, 2012 and on the new form now in use. Data on which boxes had been checked were not provided to TRAC; ICE claimed this information was not recorded in its databases.
The box(es) checked on the old form indicated at what point in the enforcement process the detainer was issued. Was ICE just initiating an investigation to determine if a person is subject to removal? Or, for example, was there a final order of removal that ICE was seeking to execute? On the new form, the phrase "initiating an investigation..." has been replaced with the statement that there "is reason to believe the individual is an alien subject to removal" with checkboxes to indicate what the reason is. The available boxes include such options as "a prior felony conviction or has been charged with a felony" or criminally convicted of "illegal entry..." For a full list of reasons, see Figure 3.
The agency's assertion that it does not record which box or boxes are checked on either the old or the replacement I-247 is surprising, if not unbelievable. If true, how does ICE management monitor whether agency policies on issuing detainers are in fact being carried out? What was the purpose of Director Morton issuing the revised directive last December if the agency's managers did not plan to assess its impact on the agents' behavior in the field? Without some way to measure what is going on, a large agency such as ICE would be acting in the dark when it attempted to manage its operations.
ICE did provide some information as to the status of the enforcement process as of March 2013, some months after the detainers were placed. These data are summarized in Table 1 and Figure 4. As of that time, a removal order had been issued for 13 percent of the individuals. For an additional 3 percent of the cases, either an arrest warrant had been issued or an immigration court proceeding had begun to secure a removal order. For all the remaining cases, an investigation was still going on (74%) or had been terminated (10%).
Information TRAC Sought on Criminal Charges
TRAC also asked for information about the specific crimes these individuals were charged with, the dates when the charges were filed, and which of them had resulted in a conviction or were still pending. In addition, TRAC requested ICE's classification of these cases by their seriousness — Level 1 being most serious to Level 3 being least serious. Because ICE has several classification systems, TRAC requested the data on each one.
- LESC Levels. The first classification system is used by Secure Communities (SC) and divides offenses into three "LESC Levels" according to their perceived seriousness. This judgment is based on the most serious charge, whether or not a conviction has yet occurred. Statistics using this system are published monthly for all "INTEROP Matches" made through the Secure Communities program.
- Detainer Threat Levels. ICE refers to a second system as the "Detainer Threat Level." Similar to LESC Levels, it takes into consideration charges that haven't yet resulted in any conviction. According to ICE, this threat level "is based upon [...] the initial crime at the time the detainer is prepared." Other distinctions between these two systems, if any, remain unclear.
- RC Threat Levels. The third system — ICE refers to this as the removal case or "RC Threat Level" — is based only on convictions. Since October 1, 2010 ICE threat levels use the prioritization of criminal offenses outlined in Director Morton's ICE Civil Immigration Enforcement Priorities memo of June 30, 2010 as amended and updated by the memorandum of the same title issued March 2, 2011.
An increasing number of local law enforcement agencies, for example, have expressed concern that SC may be sweeping up and processing for deportations individuals who were charged with only a minor violation at the cost of undermining the community's trust and cooperation with local law enforcement officers. ICE has countered that individuals currently charged with a minor violation may have committed serious crimes in the past and thereby pose a clear threat to public safety. Immigration rights advocates have challenged the wisdom of using long ago convictions to judge the current threat posed by individuals who subsequently have shown themselves to be law abiding and valuable members of society.
TRAC encountered numerous problems when it sought this information from ICE. First, without explanation, ICE failed to release any data on charges that were still pending or had not resulted in a conviction. Nor did it release information on the LESC Levels that individuals were assigned by Secure Communities. ICE gave as its reason that "LESC Levels are those reported with INTEROP Matches; they are not generally linked or reported to individual detainer cases, and accordingly [were] not being provided."
It is revealing that, while the agency initially extracted the requested data about "Detainer Threat Level" from its computerized records, the FOIA office then redacted this data. ICE FOIA officer Catrina Pavlik-Keenan, in a letter to TRAC, claimed that the release of this information "could reasonably be expected to risk circumvention of the law." Officer Pavlik-Keenan did not explain her reasoning. If the detainer threat levels reflect the priorities championed by former ICE Director Morton, who stepped down last July, it is difficult to understand the agency's claim that releasing the information would be harmful. In fact, of course, all that the requested information would reveal is whether the agency is actually carrying out its announced priorities.
Finally, ICE also withheld information on which detainers had been issued as part of the Secure Communities program and which were issued under other programs it administers. (See TRAC's discussion of these in Part I of this series.) When an agency withholds requested information, the FOIA requires it to explain the legal basis for the decision. In this case, however, it simply said it wasn't the agency's practice to provide it. On a similar note, TRAC's request for information identifying individuals who fell into any other agency priorities — "ICE fugitive, prior removal and return, entered without inspection, visa violators and overstays" — was also withheld. The agency claimed that "they cannot be [...] reliably" determined and hence were not provided.
ICE Targeting Based on Criminal Convictions
Even in the face of ICE's general foot dragging, the agency did provide case-by-case data on the most serious conviction and the date that this had occurred. It also released for each detainer case ICE's RC Threat Level — Level 1, 2, or 3 — which is based on convictions. In addition, it released information on whether or not the individual was classified as an aggravated felon. Following are the highlights from the analysis of these data.
The top 25 offenses ICE recorded as the most serious conviction for the individuals who were subject to a detainer is shown in Table 2. The much longer list covering all offenses is found in this detail table.
As noted earlier, nearly half (48%) of detainers were issued for individuals who had no recorded conviction for any criminal offense. Where there was a conviction, driving under the influence of liquor (DWI) was listed most frequently as the most serious offense. This was the case for one out of ten of the detainers. The second most frequent charge was a conviction for a simple traffic offense (6%), followed in third place by marijuana possession (2%).
Considered together, ICE data show that nearly two out of three detainers (65%) were issued for the following reasons — no charge of any kind, DWI, simple traffic offense, or marijuana possession. Next in frequency was illegal entry itself (1.9%). If the list is limited to those offenses ICE classifies as Level 1, at most less than fifty thousand or 14 percent met this standard.
Given the fact that many individuals included in these tables were already serving time in federal and state prisons at the time the detainer was placed [see Part I of this series], the low proportion who had any criminal conviction was quite unexpected.
When the detainers are classified by the seriousness of the "RC Threat Level," the data show that only 23 percent of the total — fewer than one in four — were classified at Level 1, the most serious threat. According to ICE, Level 1 is restricted to non-citizens: "convicted of 'aggravated felonies,' as defined in § 101(a)(43) of the Immigration and Nationality Act, or two or more crimes punishable by more than one year, commonly referred to as 'felonies'." This figure is larger than the 14 percent convicted of serious crimes since ICE also includes anyone convicted of two or more felonies in its Level 1 category, whatever their nature.
In nearly half of all cases the threat level recorded was "NA" — that is, the individual posed no threat since he or she had no criminal conviction. See Figure 5 and Table 3.
ICE Treats Minor Crimes As Serious Threats to Public Safety
TRAC's analysis turned up other equally troubling conditions. When the agency's threat levels based on convictions are tabulated against the most serious convictions, it turns out that minor violations can sometimes lead to the placement of an individual in the highest threat level category. This raises profound questions. Is ICE targeting the wrong individuals because the enforcement information it receives is flawed? Or are the formulas used by ICE in implementing its priorities assigning individuals incorrect threat levels?
Examples of the apparent anomalies can be seen in Table 4. This table lists the most common serious offenses that led the agency to place an individual in the highest threat category. Ranked eleventh among the most common offenses were simple traffic violations. Marijuana possession was the sixth most common. And in second place was DWI, driving while intoxicated. In each situation, the listed offense is the most serious conviction the individual had received.
In earlier studies, TRAC compared ICE's recorded threat levels with the most serious criminal convictions for individuals and found them to be quite consistent. After ICE revised its criteria, however, TRAC found that the specific offense for its highest (Level 1) seriousness category no longer appeared to match well with the offenses recorded as the most serious conviction. See section entitled "Defining 'Criminal' Activity: What ICE Counts" in TRAC's February 2012 report.
The new system, based on a combination of convictions rather than only the most serious one, is more complicated. TRAC therefore sought the detailed procedures now used by ICE to determine the RC threat levels. These detailed ICE implementation rules are contained in the document illustrated in Figure 6.
There are a number of interesting features of these operational rules:
- Aggravated felonies. Conviction for an "aggravated felony" places a person in the highest threat level under the Morton memorandum. There is a surprisingly large number of offenses in the operational rules (the right most column in Figure 6) treated as aggravated felonies that automatically place a person into Level 1. The offense categories are often broad and would appear to encompass offenses that don't necessarily qualify as "aggravated felonies." See the TRAC report discussing the definition of aggravated felonies. TRAC was able to examine the records concerning individuals who had been placed in the most serious threat level in order to determine whether or not ICE had also classified them as an aggravated felon. What we found was that for these individuals who were also classified as aggravated felons, the most serious charge recorded in the data was not limited to offenses on this list. Instead it included numerous minor crimes such as disorderly conduct, simple traffic violations, failure to appear, and illegal entry.
- Other Level 1 Rules. About half of the Level 1 offenders in the detainer data were individuals that ICE did not consider to be an aggravated felon. Here two felony convictions are required for an individual to be placed in the highest threat level. Again, however, when the most serious convictions for these cases are tabulated, one again sees numerous minor offenses as previously shown in Table 4.
- Everything not a felony is treated as equivalent. While the Morton memo referenced above said that "[s]ome misdemeanors are relatively minor and do not warrant the same degree of focus as others," the rules themselves do not make these distinctions. In addition, the rules don't include any "escape valve" so that truly minor violations aren't elevated to a higher threat level. For example, take citations punishable for at most a small fine. Or consider individuals who send in their checks to pay for traffic tickets or for walking their dog without a leash that occur in jurisdictions that define these as criminal violations (for example, the federal government). These minor transgressions are treated the same as more serious misdemeanors. Any one of them in one's criminal history results in a Level 3 classification as the default, and three of them automatically elevate the individual to a Level 2 violator — equivalent to a person who committed manslaughter or made a bomb threat.
- Reliance on recorded sentence. The rules also consider any minor offense to be a potential felony, based solely on the information recorded on the sentence, without apparently any built-in checks to require further verification when the nature of the offense and the length of sentence appear inconsistent. For example, a simple traffic violation that resulted in one day of jail time, if misinterpreted as one year instead of one day, automatically is classed as a felony. Two minor offenses with questionable sentence information automatically moves an individual into the top threat level — on par with murderers and weapons traffickers.
When the Conviction Had Occurred
Because the date of the most serious conviction was included in the information TRAC received, we were able to divide the data into whether the most serious conviction had occurred before or after the detainer was placed and analyze these records separately.
Convictions after the detainer was placed. TRAC wanted to know just how often the most serious conviction occurred after the detainer was issued, and what was the level of seriousness of that offense. Convictions that occurred after the detainer was issued — presumably relating to the arrest that triggered the detainer — would be likely to be the most serious when it was either the first conviction, or it involved a particularly serious crime. In case serious offenses took longer to resolve we divided the data by when the detainer had been placed, in order to see if a larger proportion of serious offenses turned up when a year or more had elapsed since the arrest. (Results showed that this had not occurred.)
There were 38,008 detainers where the available information recorded a conviction after the detainer was placed. Table 5 provides a breakdown of their "RC Threat Levels." Only 20 percent of these detainers had been classified as most serious (Level 1) while 58 percent were classified as least serious (Level 3). The data thus indicate that detainers are being issued in tens of thousands of cases for minor offenses where there is no prior history of serious criminal activity.
The data so far provided by the agency do not allow us to answer an important question: how often are detainers issued for a recent minor offense when the individual had been convicted of a serious crime in the past? For this TRAC would need information about the crime which had resulted in the arrest giving rise to the detainer. Unfortunately ICE has chosen not to release this information.
Convictions before the detainer was placed. TRAC also wanted to know how often long-ago convictions were a basis for placing the detainer. As shown in Table 6, about half (49%) of the most serious offenses were committed more than five years ago, and about one quarter (23%) more than 10 years ago. For more than four thousand (4%), the conviction occurred more than 20 years ago. And incredibly, there were a few where detainers involved convictions that were at least 40 years old.
Because ICE only released information on the most serious conviction, it is not possible to determine if that conviction was the only one or even the last one. But the data do show that it is common for the most serious conviction to have occurred many years prior to when ICE sought to take those individuals into custody and deport them.
As the old maxim says, actions speak louder than words. And in this case, the information about the nearly 350,000 detainers obtained and analyzed by TRAC present a reality entirely different from the one that the government has attempted to paint in its many official statements.
Footnotes  According to ICE, the FY2012 - FY2013 records include all detainers that were issued during this period, including those issued by these ICE offices and divisions: "Detention and Deportation (DDP), Fugitive Operations (FUG), Alternatives to Detention (ATD), Criminal Alien Program (CAP), Detained Docket Control (DDC), Non-Detained Docket Control (NDD), Violent Criminal Alien Section (VCS), Joint Criminal Alien Response Team (JCT), Juvenile (JUV), and Law Enforcement Area Response (LEA)). All 287(g) detainers were also included.
 The immigration "hold" is supposed to be limited to "48 hours, excluding Saturdays, Sundays, and holidays, beyond the time when the subject would otherwise been released from [...] custody" (see DHS Form I-247 and Figure 3).
 See, for example: Santa Clara County; San Francisco County; Los Angeles; Washington, D.C.; Newark; Chicago. See also: California Attorney General Kamala D. Harris' guidance on detainers and commentary by San Francisco district attorney George Gascón.
 According to the accompanying memorandum issued by ICE Director John Morton, these revisions were made to provide a better mechanism to "ensure that ICE's finite enforcement resources are dedicated, to the greatest extent possible, to individuals whose removal promotes public safety, national security, border security, and the integrity of the immigration system."
 TRAC has filed an administrative appeal contesting the failure to provide all of the data requested. The ICE appeals officer remanded the request back for reprocessing. However, the remand expressed no opinion and included no instructions on what about the agency's original response was deficient. TRAC is currently waiting for the results of this "reprocessing."
 Illegal re-entry was separately listed. Only 0.3 percent had this as their most serious charge.
 This figure includes all offenses ICE categorizes as a "Level 1" offense in its operational rules. Because TRAC did not have sentence information, broad categories like assault and larceny were all counted as Level 1 offenses, while ICE would exclude any of these where the sentence was less than one year. There were also a few offenses that ICE elevated to Level 1 based on their sentence which otherwise were classed by ICE as Level 2 or Level 3 offenses — the classification used by TRAC. All in all, while the 14 percent figure is not exact, it is likely higher than the actual number of "serious crimes" based upon ICE's definitions if full information had been available. For ICE's offense-by-offense classification rules, see Figure 6.
 For example, there are reportedly over 100,000 noncitizens serving sentences in state and federal prisons on any day (see sourcebook of criminal justice statistics). Because this statistic does not cover noncitizens serving sentences in local jails, it underestimates the actual number of individuals that ICE prison programs handle.
 When the same data are examined over time, the proportion showing any threat as well as the proportion in the Level 1 threat category both show slight declines. This could be simply a reporting artifact since classification is based upon convictions, and a greater proportion of charges may still be pending in more recent months.
 The operational rules require that there be two convictions classified as "Level 2" to move the individual into the most serious Level 1 category. There are a number of crimes listed that are classed as Level 2 if they have sentences of less than one year. Also, a Level 3 offense where the recorded sentence is one year or more is upgraded to Level 2, and two of these then rates elevation to Level 1.
 Because detainers are also placed on individuals who are currently serving a prison term to ensure that when the sentence is completed they are turned over to ICE rather than released into the community, we eliminated those records that contained an entry in a field labeled "projected release date." There were 44,735 records with projected release dates. Of these, 39,489 concerned convictions that occurred before the detainer was placed and were therefore eliminated from Table 6. This is presumably an undercount since ICE may not have recorded projected release dates for everyone who was in prison at the time.