Has Cooperation by State and Local Law Enforcement Agencies Improved ICE's Apprehension
As a complement to reforms of ICE detainer practices introduced in November 2014 by the
Priority Enforcement Program (PEP), Department of Homeland Secretary Jeh Johnson
announced a major organized outreach effort to gain back law enforcement agency
cooperation. This effort sought to counter the growing unpopularity of DHS's Secure
Communities Program. As Johnson noted: "Governors, mayors, and state and local law
enforcement officials around the country have increasingly refused to cooperate with the
[Secure Communities] program, and many have issued executive orders or signed laws
prohibiting such cooperation."
a December 22, 2015 press release, Secretary Johnson announced that this outreach effort had
been quite successful: "Of the 25 largest jurisdictions that had placed
restrictions on their own cooperation with ICE, 16 are now working with us again for the good
of public safety."
This report, the second in a two-part series, examines the data behind these claims. Detainers
sent to local law enforcement officials — often referred to as I-247s after the government form
that is used — have historically been a major method Immigration and Customs Enforcement
(ICE) employs to take into custody individuals thought to be deportable. TRAC's findings based
upon a review of individual detainer records call into question the assumed linkage between
cooperation with ICE by state and local law enforcement agencies and ICE's success in
apprehending immigrants it seeks to deport.
Figure 1. State and Local Law Enforcement Agency (LEA) Refuses ICE I-247 Detainer Request
TRAC's last report documented that ICE
still has failed to target its detainer program on noncitizens convicted of significant criminal
offenses. This report examines whether ICE actually takes individuals into custody who it
targeted with those detainers. Two conclusions emerge:
According to ICE data, DHS outreach efforts have drastically curbed reported
refusals by law enforcement agencies to transfer custody of individuals over to ICE.
During the first two months of FY 2016 ICE reports only 2 percent of the agency's I-
247 requests were refused. See Figure 1.
While recorded refusal rates fell sharply, the proportion of occasions where ICE took
custody of the individual after issuing I-247 requests has not rebounded. Indeed,
this rate has continued to decline. It is now below 40 percent. See Figure 2.
The end result is that ICE has not improved its performance through its detainer program in
apprehending individuals who the agency seeks to deport. Details for these two findings are
discussed in the sections that follow.
Figure 2. Outcome of ICE I-247 Detainer Requests
Are State and Local Law Enforcement Agencies Now Cooperating with ICE?
Under DHS's Secure Communities program, ICE data recorded a sharp rise in refusals by state
and local law enforcement agencies (LEAs). This rise in recorded refusals began in the latter
half of 2013, and peaked during the summer of 2014. Then, as shown in Figure 1, refusal rates
began falling as DHS discontinued Secure Communities and mounted an aggressive outreach
Thus, these DHS outreach efforts would appear to have drastically curbed refusals to transfer
individuals who state and local law enforcement agencies had in custody over to ICE
enforcement officers. During the first two months of FY 2016, for example, ICE data indicate
only 2 percent of ICE's I-247 requests were refused. This was down from a high of over 10
percent in the months preceding the termination of the Secure Communities program.
However, some caution is warranted before reaching any definitive conclusion. This is because
TRAC found serious quality issues in ICE's recording of refusal rates. TRAC notes that ICE
apparently does not require that information tracking law enforcement agency refusals be
systematically entered in its database. Rather, recording this information is optional. The
database field the agency uses records a variety of different reasons why ICE "lifted" — that is
withdrew — a custody transfer request.
These recorded "lift" reasons often disagree with other information ICE records on whether the
individual was actually booked into its custody. For example, ICE sometimes records an agency
refused its transfer request even though other data indicates ICE actually assumed custody of
the individual. In addition, some jurisdictions that adopted restrictive measures curtailing
cooperation with ICE's detainer program do not show up in ICE data as actually refusing ICE
It is thus unclear whether meaningful cooperation has increased, or whether ICE practices have
changed in recording when refusals occur.
Even when comparing states there seems to be a serious disconnect between recorded refusals
and whether ICE took the subject of the detainer into custody. Examining records state-by-
state for the entire FY 2014 - FY 2016 period, the failure to assume custody ranged from a high
of 86.9 percent in Indiana down to a low of 29.4 percent in Arizona. Yet both of these states
had exactly the same recorded LEA refusal rate — a mere 0.1 percent. See Table 1.
Table 1. Outcome of ICE I-247 Detainer Requests by State, FY 2014 - FY 2016*
ICE Still Failing to Take Most Detainer Targets into Custody
Since April of 2014 the majority of ICE requests to law enforcement agencies (LEAs) to assume
custody of an individual have not been effectuated. That is, examining each I-247 form that ICE
prepared, ICE assumed custody less than half of the time. Indeed, during the first two months
of FY 2016, ICE did not assume custody of more than six out of every ten individuals against
whom detainers were lodged.
These figures are in sharp contrast to ICE's success rate back in FY 2009. Then 70 percent of its
detainers resulted in ICE successfully taking the individual into custody. The drop from this 70
percent rate during FY 2009 to below 40 percent so far in FY 2016 is dramatic.
While there are potentially many legitimate reasons why ICE might issue an I-247 form but then not take
custody of the individual, the magnitude of this decline is startling.
As shown in Table 2 and Figure 2, this decline appears to have begun at a time when ICE was
rapidly expanding the number of detainers it issued under Secure Communities. It may be
that this rapid expansion exceeded ICE's capacity to actually pick up and detain the individuals
on whom it issued detainers. Only after the growth in the number of ICE detainers ended, and
indeed detainer numbers began to fall, did the proportion of those ICE took into custody
appear to stabilize at around 55 percent.
The second precipitous fall — this time from around 55 percent to below 40 percent — occurred
more recently as more and more law enforcement agencies started refusing to honor ICE
detainers. However it remains a puzzle — if in fact DHS's outreach efforts have improved LEA
cooperation as Secretary Johnson reports — why no corresponding increase has occurred in the
occasions where individuals were transferred to ICE custody. Using this latter criteria,
Homeland Security's outreach to state and local law enforcement agencies would not appear as
yet to have brought about any major impact on ICE's success in taking individuals thought to be
deportable into custody.
Details Now Available for Individual State and Local Law Enforcement Agencies
TRAC has just released a new free detainer app that
allows users to examine the details month-by-month on ICE I-247 forms sent to each
individual state and local law enforcement agency during FY 2003 to FY 2016. Available are
ICE records on whether the detainer request
was refused, as well as whether ICE successfully took the individual sought into custody. Users
can drill in to analyze what happened to detainers depending upon whether the individual had
ever been convicted of any crime, as well as drill further down to the specific criminal offense -
drunk driving versus homicide, for example. Demographic information by age, citizenship, and
gender can also be examined.
The data show that while Texas leads the country in receiving the largest number of ICE I-247s since FY 2014,
California law enforcement agencies have racked up the largest number of recorded refusals of
ICE detainers. That state also had the largest number of occasions in which ICE fails to assume
custody of an individual after it issued an I-247.
Table 3 presents some highlights from this wealth of data. If the proportion of detainer
requests that state and local LEAs refused to honor is examined during the months immediately
prior to the ending of Secure Communities, California LEAs not surprisingly held each of the top
spots. The Santa Clara County Main Jail had the highest recorded refusal rate in the nation at
88.2 percent. However, in the latest available data since Secretary Johnson's reforms were
implemented (July - November 2015), the recorded refusal rate has plummeted to 4.8 percent.
Yet ICE still failed to assume custody of virtually every individual on whom it sent an I-247 form
to the Santa Clara authorities both before and after the outreach effort by DHS occurred. This
raises the question: Did Santa Clara's cooperation actually increase? Or did ICE simply stop
recording refusals that occurred? Or is there some other explanation for these wildly dissimilar
A similar divergent picture emerges, depending upon the measure of cooperation, for the LEAs
who had the second and third highest recorded refusal rates in the months immediately prior
to DHS's outreach efforts. The Santa Rita Jail in Alameda County, California, was second with an
80.1 percent recorded refusal rate, while the San Mateo County Jail in California had 74.1
percent. Each of these saw their recorded refusal rates plummet to 0.9 percent and 2.4
percent, respectively, in the most recent ICE records. Yet ICE still failed to take into custody
the vast majority of individuals held in these jails during this same period on whom I-247s were
However, recorded LEA refusal rates did not always plummet between June 2014 and
November 2015. The fourth highest recorded refusal rate in the nation of 68.7 before PEP was
announced was by the Ventura County Jail in California. After the reforms, ICE recorded that
the refusal rate declined. However, this facility still refused half of all I-247s it received. ICE
data also showed a parallel increase in the rate individuals held in Ventura County Jail were
taken into ICE custody — from 0.4 percent before the reforms to 12.5 percent after they
Beyond California, three other law enforcement agencies were among the top ten recording
the highest refusal rates before PEP was announced. These were the Philadelphia Detention
facility in Pennsylvania (sixth place), the King County Adult Jail in Washington State (eighth
place), and the Cook County Jail in Illinois (tenth place) — all facilities located in major
metropolitan areas in the nation. King County and Cook County facilities refusal rates each
plummeted in the latest ICE data, while ICE failed to show any increase in the proportion of
individuals with I-247s taken into custody. The Philadelphia Detention facility, in contrast,
showed only a modest drop in recorded refusal rates (from 67.3 to 50.0 percent) and a similar
small increase in those ICE took into custody (from 14.5 to 25 percent).
Similar data for additional facilities can be found below in Table 3. Information for other
specific time periods can be found in TRAC's new free detainer app.
enforcement agencies by county that passed legislation or adopted restrictive policies limiting their
cooperation in turning over individuals to ICE custody in response to detainers the agency issued.
The data for the most recent period reflect ICE's records as of April 4, 2016. While ICE data indicate that the
typical lag between when an I-247 is issued and the transfer to ICE custody takes place is 19 days and the most
common is just a single day, a small number of recorded custody transfers occur more than four months after a I-
247 form is issued. Thus, for the most recent time period these figures may slightly undercount the transfers that
eventually will occur.