Get Informed

Subscribe to our newsletters for regular updates, analysis and context straight to your email.

Close Newsletter Signup

New Documents Reveal How ICE Mines Local Police Databases Across the Country

New Documents Reveal How ICE Mines Local Police Databases Across the Country


In cities across the country, Immigration and Customs Enforcement Homeland Security Investigations agents can mine local police reports using COPLINK, a data program little known outside law enforcement circles. While public records have revealed ICE’s access to this program in the past, new documents, obtained by the ACLU of Massachusetts and shared with The Appeal, offer the first up-close glimpse at how the program allows ICE to access millions of sensitive police records.

The software ingests local police databases, allowing users to map out people’s social networks and browse data that could include their countries of origin, license plate numbers, home addresses, alleged gang membership records, and more.

ICE HSI agents have direct access to the Massachusetts version of the COPLINK system, which receives records from Massachusetts’s Registry of Motor Vehicles, Board of Probation, and at least local 25 police agencies. As NPR has previously reported, ICE also has direct access to COPLINK-powered databases in other jurisdictions across the country, gathering data from dozens of police departments across Los Angeles County and Arizona.

Law enforcement officials and former ICE agents say the sharing of these databases and analytic tools helps ICE Homeland Security Investigations — the agency’s investigative arm — tackle serious crimes, like child pornographyand money laundering. But immigration advocates point out that ICE HSI agents are also involved in questionable immigration enforcement actions nationwide. These include investigations into undocumented residents accused of gang membership, sometimes based on flimsy evidence, and controversial workplace raids, which acting ICE chief Tom Homan recently announced plans to quadruple.

Earlier this month, at a slaughterhouse in Bean Station, Tennessee, ICE HSI agents carried out the largest workplace raid in over a decade, rounding upnearly 100 immigrants. Advocates say this raid signals a return to George W. Bush-era style workplace enforcement actions, which swept up hundreds of immigrants in towns like New Bedford, Massachusetts, over a decade ago.

Department of Homeland Security (DHS) data-sharing agreements obtained by The Appeal through the Freedom of Information Act make clear that ICE agents can access COPLINK “in the same manner” as local law enforcement for immigration enforcement purposes.

The constantly updated police records in COPLINK, arising from day-to-day police encounters, can be indispensable for ICE HSI agents, who often need to find addresses, cars, phone numbers, and associates that are not necessarily housed in federal or private sector databases, according to Massachusetts police and former ICE agents interviewed by The Appeal. They can help ICE officers conduct background research on employees before a workplace enforcement action or when planning logistics for a gang raid.

In a phone call with The Appeal, ICE spokesperson John Mohan declined to discuss ICE’s use of COPLINK. “We don’t talk about the techniques or tools [ICE agents] use,” said Mohan. The Appeal has since filed a Freedom of Information request for data on ICE agents’ querying of COPLINK databases across the country.

Notably, many of the major police departments that feed into COPLINK, and allow ICE direct access to their data, are situated in so-called “sanctuary” jurisdictions that have promised some measure of protection to undocumented immigrants. Los AngelesSan Gabriel, and Pasadena police, for example, feed into a Los Angeles County-wide COPLINK database, and Boston, Somerville, and Cambridge police feed into Massachusetts’s COPLINK regional database.

Kade Crockford, director of the ACLU of Massachusetts’s Technology for Liberty Program, argues that although it’s unclear whether sanctuary cities can legally withhold data from ICE, they should not actively help ICE agents by giving them direct access to local systems like COPLINK.

“By giving ICE and other federal agencies credentialed access to state and local law enforcement databases, communities are perhaps unintentionally endangering immigrants and other groups who may be targeted by the federal government,” Crockford said in a statement to The Appeal. “Data is toxic, and local communities must understand exactly what their police are collecting, how they are retaining it, and under what circumstances outside entities like ICE can access it.”

How the Program Works

COPLINK was initially developed by University of Arizona researchers in collaboration with the Tucson Police Department in 1998, as a way to share information between local police departments regionally. It has since spread to over 5,100 law enforcement jurisdictions throughout the United States and is now owned by the technology firm Forensic Logic.

ICE’s access to local COPLINK police databases has grown as well. In 2008, ICE signed data-sharing agreements with AZLINK, the COPLINK data hub in Arizona, and with IRIS, the COPLINK hub in Los Angeles County. It is unclear when ICE first formally gained access to Massachusetts police data through COPLINK, but a training roster obtained from the Massachusetts State Police listed numerous ICE agents and analysts next to dates, which correspond to COPLINK training sessions in 2015 and 2016.

In Massachusetts’s version of COPLINK, 25 police agencies across the state automatically feed almost all the data from their records management systems — including arrests, complaints, and citation reports — into COPLINK, according to the documents. Databases from another 13 police departments, in addition to the state’s Department of Correction, Parole Board, and sex offender registry, were being integrated into the program as of last year. According to Lieutenant Colonel Dermott Quinn, commander of the Massachusetts State Police’s Division of Investigative Services, the system also intakes accident reports, parking tickets, and field interview notes from local police departments.

ICE HSI agents have virtually unfettered access to COPLINK in Massachusetts. Similarly, ICE’s data-sharing agreements with Arizona’s AZLINK and the Los Angeles Sheriff’s Department’s IRIS, both COPLINK systems, state explicitly that ICE agents are free to mine their data. In 2011, for example, Department of Homeland Security users, including ICE personnel, searched AZLINK, an Arizona regional data system that uses COPLINK software, thousands of times over a six-month period.

Drawing from this body of records, COPLINK allows users to search for individuals, organizations, and vehicles, among other things. These searches can be narrowed down using filters, such as a person’s race, hair color, eye color, complexion, ethnicity, and country of origin. The slide below also shows that users can search for people based on their residence and physical marks on their bodies.

In the slide below, the state police attempted to redact information, concealing their investigative techniques with black boxes. The black boxes on the files, however, can be deleted. The Massachusetts State Police did not respond to The Appeal’s inquiry about the original redactions. The Appeal has chosen to publish select portions of the redacted documents, with some new redactions added to protect people’s identities, to show the tremendous search capabilities the program affords users, whether they be local cops or ICE agents. This document, for example, shows that police officers and ICE agents using COPLINK can also filter through masses of records to find people alleged to be gang members.

Work that would have taken months in the past, and required piecing together disparate data points from agencies across the state, can now be done with a few clicks, says Lieutenant Michael Kmiec of the Lynn Police Department in Massachusetts. “It speeds up the process in looking at information in other departments. Instead of calling the other department, you can pull it directly,” said Lieutenant Kmiec to this reporter, then a freelancer with ProPublica, in a phone call last September. “Regarding the FBI and ICE, if they’re looking to look up a name or vehicle information, it’s a database accessible to them to look up things quickly.”

Clashing Interests: Law Enforcement vs. Immigrants’ Rights

Law enforcement officials interviewed by The Appeal argue the sharing of residents’ data with ICE has clear benefits at both local and federal levels.

Local databases and analytic tools, like Massachusetts’s COPLINK system, are a treasure trove for ICE HSI special agents because they provide granular data that federal agents often lack, says John Sandweg, former acting director of ICE in 2013 and 2014. “Say you have a suspect and you’re working on a federal case, you want to tap into the information, do associate mapping, link analysis — it can be huge for an investigation,” said Sandweg in a phone call. “They’re going to have more information than the feds.”

For ICE’s worksite enforcement operations, analysts rely on this type of local police data to do deeper background checks on undocumented employees. Claude Arnold, a retired ICE special agent who led the agency’s national gang initiative, Operation Community Shield, explained how it works. “Say they’re preparing a worksite case and they’re running all the employees of the company. They’re going to run criminal checks, but the NCIC [National Crime Information Center] and other federal databases wouldn’t have field interview cards [in which police transcribe notes on interactions with or observations of civilians],” he told The Appeal. While NCIC contains some gang information, COPLINK would be more likely to have information on employers, associates, and hangout spots, he explained. “It would sweeten the pot if a bunch of the employees are criminal aliens. When you go to the U.S. Attorney’s Office, you can say not only are they employed illegally, there are two gang bangers and this guy has two illegal entries.”

The help goes both ways. Local law enforcement sometimes calls on ICE HSI to deport alleged gang members when local authorities don’t have enough evidence to lead successful investigations themselves, Sandweg points out“Say you got a guy who’s been deported and he’s back, it may be difficult to find probable cause on him for a criminal investigation, but it’s easy to do civil proceedings for a deportation,” said Sandweg. “Maybe some innocent people have been picked up, but HSI has got a lot of violent felons off the streets.”

Arnold reiterated this point, noting that local police routinely rely on this kind of collaboration with ICE. “Typically, what would happen is the local police would give me a list of the people in that gang, and then we just do a full workup on all of them, and that’s what you’d use COPLINK for,” said Arnold, referring to research done on the targets’ immigration status and their potential criminal and civil immigration violations.

Lieutenant Quinn of the Massachusetts State Police says that people should not be concerned about these search capabilities being shared with ICE Homeland Security Investigations because much of HSI’s work focuses on high-level cases.

“We work with HSI on the criminal side. They do a lot of criminal work and that’s what we help with,” said Quinn to this reporter, then a freelancer with ProPublica. “It’s just plain old criminal work that has been enhanced by these tools,” he added, referring to the fact that ICE HSI carries out a wide range of investigations, from financial and cyber crimes to export enforcement and weapons smuggling.

But those aren’t the only types of investigations handled by ICE HSI, explains Maddie Thomson, an immigration attorney at Boston’s Community Law Office. The example she cites isn’t about a COPLINK case, but does show the breadth of the agents’ work and their reliance on local police records.

In March, one of Thomson’s clients, a young Salvadoran man who was classified as a gang member despite having no criminal record, was deported after having entered the country without inspection, a civil matter, following an ICE HSI investigation.

The evidence cited in immigration court for his alleged membership in MS-13 were all local police documents, school police observations and field interrogation and observation (FIO) records, which ICE in this case pulled from a Boston police gang database. In smaller jurisdictions that do not feed records into a gang database, however, ICE could have just as easily used COPLINK to pull the same records.

In immigration court, Thomson’s client’s familiarity with other alleged gang members helped greenlight his deportation. “The DHS attorney would ask him, ‘Do you know this person? Do you know that person is a gang member?’ and he’d say, ‘No, I just knew him as someone playing pick up soccer in the park,’” recalled Thomson. In February 2018, her client received a final removal order. He was put on a plane to El Salvador last month.

ICE’s willingness to remove undocumented people who could not be prosecuted through domestic criminal courts violates the spirit of due process, argues Sarah Sherman Stokes, a clinical instructor in the Immigrants’ Rights Clinic at Boston University School of Law.

“For them, it’s better if a bunch of innocent people get apprehended than letting one guilty person go free,” said Stokes, who has represented young immigrants put into deportation proceedings due to ICE HSI investigations. “I’m incredibly troubled by the kind of information-sharing happening with local law enforcement and ICE, and am increasingly disturbed by the role of HSI in apprehending immigrants, whom there otherwise wouldn’t be probable cause for law enforcement to apprehend.” ICE declined to comment on HSI’s role in immigration proceedings.

Where and How Local Police Data Is Being Shared

Several of the 25 police departments sharing data with COPLINK are in the greater Boston area, home to many of the state’s largest and densest undocumented communities, even so-called “sanctuary” cities, such as Bostonand Cambridge.

In the interactive map below, viewers can see where local police departments are fully sharing or in the process of sharing data with COPLINK, and, in effect, with ICE. Click on the department dots to see their data-sharing policies and on the census tracts to see the density of foreign-born, non-citizen residents, as estimated by a 2016 U.S. census dataset:

ICE Access to Local Police Data

Map created by georgejosephmapping in CARTO

How an individual is defined as a gang member varies based on local policing practices, but, as advocates point out, these important distinctions are blurred in an aggregated data system and can net the wrong people into COPLINK.

The software ingests police data from local records management databases, but the definition of gang membership, for example, varies across the state, says Captain Harry Hess, a technical services commander at the Fitchburg Police Department. In Fitchburg, Hess said in a phone interview with The Appeal, officers base gang classification on a case-by-case assessment of things like tattoos, interviews with the individual in question, or their known associates.

In Boston, on the other hand, police use a more defined points-based formula, as Crockford recently explained in a piece for The Appeal. This points-classification system is based on adding up questionable “gang” indicators, such as who an individual has been seen speaking to, their clothing colors, and police judgements about their photos on Facebook.

But that means COPLINK users, whether they are cops from neighboring counties or federal immigration agents, have no way of knowing what local police practices culminated in an individual getting labeled as a gang member.

“If you’re looking for a suspect in a car break-in and you knew he was a Latin King in a certain age range, you want to have those labels to search for a suspect that might match,” said Kristi Fritscher, a crime analyst at the Fitchburg Police Department in a phone interview with The Appeal. “But how they came to that conclusion is not in there.”

The varying data-collection practices used by departments feeding into COPLINK could raise a “junk in, junk out” problem for immigrants or anyone added into the system, Crockford argues.

“If someone from ICE HSI searched for gang members in Massachusetts, that person doesn’t necessarily know if someone has been identified as a gang member by a Somerville police official just because they had a hunch or by a Waltham officer because they killed three people,” said Crockford. “There’s a flattening that takes place, where allegations by local police who may have no idea what they’re talking about, suddenly looks the same as allegations of real violence.”

Sanctuary Cities Respond

Regardless of the quality of the information, immigration advocates say, it’s dangerous for cities to share it with ICE and disappointing that sanctuary cities would do so.

“The fact that police are sharing information [with ICE] clearly violates the spirit of the sanctuary idea,” said Thomson, referring to COPLINK. “So it’s very easy for cities to get liberal cred, while also participating in these systems and actively helping to deport people.”

Jonathan Fu, a spokesperson for the city of San Gabriel, California, responded to a request for comment from The Appeal on why his sanctuary city uses COPLINK. “The San Gabriel Police Department has access to COPLINK, which allows us to search for data related to criminal reports. The system does not allow us to list immigration status, nor do we provide such information to COPLINK.” Fu added that the city wants undocumented immigrants to be able to come forward with relevant information about crimes to law enforcement without fear of deportation, but he did not bring up any plans to alter the department’s data-sharing practices.

Denise Taylor, a spokesperson for the city of Somerville, Massachusetts, responded to a similar request from The Appeal. “The safety of all of our residents and the protection of our families regardless of status remains a top priority for the city and undergirds our sanctuary city/safe city policies,” she wrote in an email. “[We] will certainly carefully review this with local law enforcement and other sanctuary cities to determine the best policy going forward.”

The mayors’ offices of the two other Massachusetts’s sanctuary cities that feed into COPLINK — Boston and Cambridge — declined to comment on their cities’ use of the program. Mayors’ offices in two other California sanctuary cities that contribute to a Los Angeles County-wide COPLINK database, Los Angeles and Pasadena, did not comment on the matter.

Other elected officials in those cities interviewed by The Appeal expressed varying degrees of discomfort with this data-sharing.

Lydia Edwards, a Boston city council member and attorney from the heavily immigrant community of East Boston, noted that currently, Boston’s sanctuary policy only shields undocumented immigrants from ICE’s civil detainer requests to hold arrestees in jail for pick-up by ICE agents, not from information-sharing more broadly.

In a statement, Edwards noted that while this detainer component is beneficial, it’s not enough. “As a city, we should be transparent about the data we collect, who we are sharing it with, and what public purpose it serves so that these agreements can be fully vetted,” Edwards said.

Likewise, J.T. Scott, an alderman from Somerville, said in a statement that the Somerville police department’s participation in COPLINK “certainly raises troubling questions about how much information ICE HSI can obtain about residents of our city,” and added that he will be raising the issue through the city’s Board of Aldermen.

On the other hand, Quinton Zondervan, a Cambridge city council member, argued that data-sharing with ICE HSI is necessary for serious criminal investigations, even if it means ICE can use the information for immigration-related purposes. “It’s very hard to prevent that,” said Zondervan in a phone call with The Appeal.

While some local officials are calling for greater transparency regarding their data-sharing with ICE, none have definitively called for their police departments to stop using COPLINK or limit ICE’s access to their residents’ data through, for example, amendments to the agencies’ data-sharing agreements. While legal experts differ on whether or not cities could prevent ICE from accessing local police data altogether, Crockford says amending such agreements would weaken ICE’s ability to mine millions of records at once and instantly piece together people’s connections.

Localities seeking to protect immigrant communities should “conduct comprehensive reviews of their information sharing systems,” said Crockford. “That’s just as important as refusing to honor detainer requests, but because all of these police data systems were built and maintained out of the public view, it hasn’t gotten the same attention.”

Thomson said it’s hard to get lawmakers to take action on this data-sharing issue given undocumented immigrants’ current lack of political power.

“Generally, this is a population whose experiences are not shared by voters,” said Thomson, since most voters do not have families of mixed legal status. “So it is really easy for elected officials to not really address what’s happening.”

The Epidemic of Brady Violations: Explained

Illustration by Hisashi Okawa

The Epidemic of Brady Violations: Explained


In our Explainer series, Justice Collaborative lawyers and other legal experts help unpack some of the most complicated issues in the criminal justice system. We break down the problems behind the headlines—like bail, civil asset forfeiture, or the Brady doctrine—so that everyone can understand them. Wherever possible, we try to utilize the stories of those affected by the criminal justice system to show how these laws and principles should work, and how they often fail. We will update our Explainers quarterly to keep them current.

In 2003, Brooklyn prosecutors charged John Giuca with killing a 19-year-old college student to earn “street cred” with the Ghetto Mafia gang. A jailhouse informant’s testimony claiming that Giuca confessed to him in prison formed a critical piece of the government’s case. At trial, the government insisted that the jailhouse snitch had no deal with the government. That was a lie. Fifteen years later, a court of appeals tossed out Giuca’s murder conviction, finding that the jury may well have acquitted Giuca if it had known the truth.

In courtrooms across America, prosecutors regularly withhold evidence from the defense that could blow holes in their cases. That’s a violation of the Brady doctrine, based on a 1963 Supreme Court ruling that requires them to disclose any information favorable to the defense.

When prosecutors withhold evidence they are duty-bound to turn over, they undermine the Constitution, the Supreme Court’s case law, and the premise of justice.

What is the Brady doctrine?

Under the Supreme Court’s 1963 case Brady v. Maryland, the prosecution has an ironclad duty to disclose, before trial, evidence that could undercut its case — “Brady material.” If the prosecution doesn’t do that, it violates the Constitution. The case involved John Leo Brady, a man convicted of first-degree murder because prosecutors suppressed evidence that his accomplice confessed to the actual killing. The Supreme Court recognized that “[s]ociety wins not only when the guilty are convicted but when criminal trials are fair,” and that a prosecutor should not be the “architect of a proceeding that does not comport with standards of justice.”

What qualifies as Brady material?

Brady material includes any evidence that is favorable to the defense — and that encompasses a lot of information. That means anything that can help the defense attack the prosecution’s case.

The prosecutor is responsible for disclosing anything known by members of the prosecution team , which includes law enforcement, forensic investigators, and other experts. Under Kyles v. Whitley, a case interpreting the Brady doctrine, the government cannot claim ignorance. It must actually find out what information is in the files of the people on whose work and expertise it relies. This rule is important. Making prosecutors immune from Brady when the material is in the police or analysts’ files would create perverse incentives for prosecutors not to know about information favorable to the defense.

1. Evidence suggesting someone other than the accused committed the crime

In a case that landed before the U.S. Supreme Court during its 2016–17 term, Turner v. United States, for example, prosecutors withheld evidence during a 1985 trial that a witness saw someone other than the defendant lurking around the crime scene, and then fleeing from it, shortly after the murder. That person had a lengthy criminal record, which included committing violent crimes in that part of town. The prosecution never turned over that information, which was clearly Brady material. [Thomas L Dybdahl / The Guardian & Marshall Project]

In a 6–2 opinion, the U.S. Supreme Court later affirmed the conviction, taking comfort in the prosecutor’s claim that they were now carefully guarding their own hen house and had adopted a “generous policy of discovery.” [Jessica Brand / Slate]

2. Evidence of a witness’s prior inconsistent statements (known as Giglio material)

In 1985, police recovered Bridgett Lamon’s body in a dumpster in east Anaheim, California. They also recovered semen from her vagina. Three months later, police found Catherine Tameny dead in her Anaheim apartment, with semen deposited on her underwear in the bathroom and saliva on her chest. Two decades later, after cold hits revealed various suspects, Lynn Johnson went to trial for the Lamon murder — he matched the semen, and Wendell Lemond went to trial for the Tameny murder — he did not match the semen. Both men were convicted.

Forensic analyst Mary Hong gave critical testimony in both cases, but her testimony could not be reconciled. In the Lamon case, she claimed that the low quantity of semen found meant it had been left there within 24 hours of the crime scene, which made Johnson the likely killer. But 15 months later, in the Tameny case, she testified that the low quantity of semen left on the underwear meant it was deposited more than 24 hours prior to the crime, and so the jury should not worry about the lack of a match with Lemond. Tameny could have had sex with someone else but then, over a day later, Lemond could have murdered her.

The government never disclosed Hong’s inconsistent testimony, but years later, a public defender discovered it. “It’s beyond sick. [In] how many other cases has [Hong] adjusted her opinion so it could work for the prosecution?” the public defender asked. Now, a judge has ordered a briefing on whether the government failed to turn over material Brady evidence — evidence of the examiner’s inconsistent analysis, or, put differently, its flip-flopping. [R. Scott Moxley / OC Weekly]

3. Evidence of a witness’s motive to lie (Banks v. Dretke)

Witnesses don’t always testify out of the goodness of their hearts, and any reason they might have to gain favor with the prosecution is Brady material. In a 1992 trial, the state of Texas accused Cameron Todd Willingham of killing his children by arson, eventually putting him to death. Not only did the state use now-discredited arson evidence, but the prosecutor also failed to disclose a deal the prosecutor struck with a jailhouse snitch so the snitch could receive a reduction in charges. This should have been disclosed as Brady material. [Jordan Smith / The Intercept] [Radley Balko / Washington Post] [David Grann / The New Yorker]

4. Information that casts doubt on the credibility of the police

In Baltimore, video uncovered during the summer of 2017 showed what appeared to be police officers planting evidence at a crime scene (they later claimed it was a “reenactment”). That evidence cast serious doubt on the officers’ credibility and on the integrity of their investigation, and therefore qualified as Brady material. [Eric Levenson, Lauren del Valle & Darran Simon / CNN]

5. Information that casts doubt on a crime lab technician

In San Diego, California, a judge recently reversed Florencio Jose Dominguez’s murder conviction. Four days before trial, the San Diego Police Department Crime Lab changed how it interpreted DNA mixtures — biological evidence containing DNA from more than one person. Under the new DNA interpretation protocols, analysts could draw no conclusions about whether Dominguez contributed to the DNA profile. And yet the DNA analyst — who spearheaded the change — took the stand and testified using the old procedures and declared a match between Dominguez’s known profile and that found at the crime scene without telling the defense about the change in protocol. The defense lawyer learned about it by accident after Dominguez spent five years behind bars. In October 2017, a court reversed Dominguez’s conviction, finding a Brady violation. [Greg Moran / San Diego Union-Tribune]

In Massachusetts, for years, state lab chemist Sonja Farak used many of the drugs she was supposed to be testing, and she did so while at work. Did that compromise the integrity of her work and investigation? You bet. That evidence was therefore Brady material in all the cases where she testified as an expert witness. (The prosecution, however, failed to timely disclose it, which a judge later called a “fraud on the court.”) [Jennifer Laurin / In Justice Today] [Jessica Brand / In Justice Today]

As a general rule: If the defense would want to know about it, it’s probably Brady.

The Effects of Brady Violations Are Serious

Brady violations not only send potentially innocent people to prison, but they reinforce a win-at-all costs mentality that undermines the pursuit of justice.

John Thompson spent 18 years in prison, 14 on death row, convicted of a murder and armed robbery. The Orleans Parish (LA) district attorney’s office suppressed blood evidence that exculpated him, discovered by the defense just before Thompson’s execution date. Released in 2003, he died of a heart attack in 2017, his life in the free world cut short. [Radley Balko / Washington Post]

Michael Wearry was also placed on Louisiana’s death row for a 1998 murder because of the prosecution’s failure to play fair. Their case relied on two eyewitness accounts. After a jury convicted Wearry and sentenced him to death, significant Brady evidence emerged. Contrary to the prosecution’s representations at trial, one of the witnesses had twice sought a more lenient sentence in exchange for his testimony, and the police had told him they would “talk to the DA if he told the truth.” Evidence also showed that the other witness gave an account that was impossible. He accused an accomplice of running, bending, lifting, and crawling into a cargo space. But the prosecutor failed to disclose that the guy had just had knee surgery and couldn’t do any of that. The state also kept mum about a prison informant’s attempt to manipulate other witnesses into giving incriminating statements. In 2016, the Supreme Court reversed Wearry’s death sentence and conviction. [Heidi Kinchen / The Advocate]

And there is the aforementioned Cameron Todd Willingham. As has been well documented, he ended up on death row after prosecutors presented shoddy arson evidence as fact and failed to disclose evidence that its star witness made a deal with the government — the most basic of Brady evidence. In February 2004, the state of Texas executed him. [Maurice Possley / Washington Post]

Why do Brady violations persist?

Ask any public defender in the country, and they will tell you that Brady violations occur regularly in the courthouse. The National Registry of Exonerations estimates that over 50 percent of wrongful convictions occur because of official misconduct.

At best, prosecutors commit Brady violations because they are fallible, and they suffer from confirmation bias, which leads them to focus on evidence that validates what they already believe. At worst, they care only about conviction rates, and, as former Ninth Circuit appellate judge Alex Kozinski believes, “they consider [Brady violations] feathers in their caps.” [Thomas Dybdahl / In Justice Today]

Meanwhile, there are surely cases where evidence is never uncovered, never heard by a judge or a jury, and the defendant remains unaware of its existence. Even when it is uncovered, however, the penalties for prosecutors can be virtually meaningless. Here’s why:

6. Because prosecutors claim evidence is not “material”

Prosecutors often argue that they only must turn over exculpatory evidence that is “material” — evidence that creates a reasonable probability of a different outcome. There are several problems with this argument.

First, the government is relying on a post-trial standard. When the defense uncovers suppressed evidence after trial, it must show that there is a reasonable probability of a different outcome to obtain a new trial (discussed more later). But the Supreme Court has made clear that while a failure to disclose may only rise to a constitutional violation if the evidence is material, the government’s pretrial disclosure obligations are broader. (Strickler v. Greene, see also the D.C. Court of Appeals in Boyd v. United States)

This makes sense because prosecutors, convinced they have a strong case and that the defendant is guilty, aren’t very good at assessing what evidence might be “material” to the defense. For this reason, the Supreme Court has urged prosecutors to err on the side of disclosure. (Kyles v. Whitley)

7. Because courts are hesitant to reverse convictions

Courts will only reverse a conviction and send it back for a new trial based on a Brady violation if the suppressed evidence was “material” — if there is a reasonable probability that, had it been disclosed, it could have affected the case’s outcome. There must be a chance, in other words, that confronted with the evidence, at least one juror might have found reasonable doubt. [Lorenzo Johnson / HuffPost]

But courts regularly apply this standard in the strictest way possible. Mario Owens, one of three men on Colorado’s death row, has claimed that prosecutors committed at least 22 Brady violations in his case, including paying witnesses thousands of dollars to testify against him. According to the defense complaint, prosecutors also threatened one of the main witnesses with murder charges if he didn’t testify against Owens. Numerous witnesses received lenient sentences. The prosecutor’s office didn’t dispute this misconduct, but instead argued that evidence of these extensive payments and leniency would not have mattered to the case’s outcome. A judge who heard the claims agreed, and the case will now go up on appeal. [Susan Greene / The Colorado Independent]

8. It is nearly impossible to sue prosecutors in civil court.

Individual prosecutors are shielded by absolute immunity for “prosecutorial actions.” They have qualified immunity for “administrative” or “investigative” acts — still an exceedingly difficult standard to meet that requires showing a violation of a clearly established statutory or constitutional right. [Alexa Van Brunt / The Guardian] [Will Baude / Washington Post]

Notably, a petition for certiorari filed in January asks the Supreme Court to reconsider its qualified immunity doctrine. In that case, police officers choked a woman to death while trying to get drugs out of her mouth as she sat in the back of the police car. Showing how hard it is to sue prosecutors or the police, the Fifth Circuit found immunity: “Previous law has provided no guidance regarding what is precisely reasonable and what is unreasonable regarding the use of force to an individual’s throat where the individual appears to be concealing something in their mouth.” [Will Baude / Washington Post]

9. State bar associations rarely discipline bad actors.

In New Orleans, there is a long-standing culture of Brady violations. As of 2015, courts had overturned at least 36 convictions due to prosecutors behaving badly in New Orleans — nine of them involved defendants on death row. [Radley Balko / Washington Post] According to Radley Balko’s investigative report, a defense attorney filed eight complaints with the Office of Disciplinary Council in 2011. It took three and a half years for him to receive notice that the office even received them. [Radley Balko / Washington Post]

Other studies confirm that state bars rarely discipline Brady violators. In 2010, for example, the Northern California Innocence Project found that the state bar “publicly disciplined only one percent of the prosecutors in the 600 in cases in which the courts found prosecutorial misconduct.” [Bidish Sarma / American Constitution Society blog]

10. Courts are hesitant to impose disciplinary sanctions.

Recently, the Louisiana Supreme Court gave a Vernon Parish assistant district attorney a pass after he explicitly told a detective to remove exculpatory evidence from a police report before turning it over to the defense, and also instructed the trial prosecutor not to turn the information over to the defense. The real police report only came to light during the defendant’s third trial. The Louisiana Supreme Court cleared the prosecutor of misconduct because the evidence came to light during the third trial, and therefore did not affect the case’s outcome. [John Simerman / The New Orleans Advocate]

11. Prosecutors can avoid admitting misconduct by negotiating deals.

In February 2018, the state of Missouri and the city of St. Louis settled a civil rights lawsuit brought by the family of George Allen Jr., who was wrongfully convicted of killing and raping a woman in 1982. Allen was sentenced to 95 years in prison when he was 26 years old and served nearly 30 years before his conviction was thrown out in 2012 due to Brady violations and new DNA evidence. The state and city defendants would not admit to claims of withholding Brady evidence but agreed to pay nearly $14 million to Allen’s family. [Rachel Lippmann / St. Louis Public Radio]

In New Orleans, the government recently agreed to a reduction in sentence from life without parole to 25 years for Jeremy Burse, accused of shooting a friend during a botched robbery attempt when he was 15. Prosecuting attorney Laura Rodrigue claimed she did this because of the “devastating losses” both families suffered, and that it was part of the office’s “restorative justice program.” But in reality, settling the case helped prevent a finding about whether she threatened to prosecute the state’s key witness if he didn’t testify, and then offered him a new attorney for his pending case in exchange for testimony — evidence not disclosed at trial. The witness later wrote that he had not seen the shooting, but Rodrigue “said that they would charge me for having a role in [the] murder if I did not testify. I was very scared.” [Victoria Law / In Justice Today]

12. And defending bad behavior is not necessarily a hindrance to political advancement.

President Trump nominated Kyle Duncan to the Fifth Circuit Court of Appeals. Duncan defended the prosecutors who sent John Thompson, an innocent man, to Louisiana’s death row by suppressing blood evidence that exculpated him. Before the Supreme Court, Duncan successfully argued that the office should bear no liability for actions that kept Thompson in prison for 18 years. [Laverne Thompson / New York Times]

Some places have rules that protect bad actors — especially when they are police. In Los Angeles, the sheriff’s department has a list of 300 deputies who have a history of lying and misconduct — evidence that could undermine their credibility on the stand. Some of these individuals are accused of lying to cover up their own illegal use of force. But the list is secret. Even prosecutors don’t have access to it. The Supreme Court of California is set to review whether this is constitutional in 2018, which could have ripple effects across the state, where similar police protections apply. [Maya Lau, Ben Poston, & Corina Knoll / Los Angeles Times]

Efforts to curb Brady violations

In several cases, defendants have successfully challenged Brady violations and won their freedom. And occasionally, the prosecutors who failed to disclose information have faced repercussions.

In Virginia in 2011, a U.S. District Court overturned Justin Wolfe’s capital murder conviction after the prosecution’s key witness revealed that police strong-armed him into testifying and fed him information about the case — again, Brady evidence that should have been disclosed. [Dahlia Lithwick / Slate]

In New Orleans, Robert Jones is currently suing the city after he spent 23 years in jail for a rape and murder he did not commit. He alleges that since the 1970s, NOLA prosecutors have buried evidence favorable to the defense in at least 45 prosecutions. That includes his case, where prosecutors withheld evidence strongly indicating that a different man, Lester Jones (no relation), committed the rape and murder. The victim’s description of the attacker matched Lester; the rape took place near Lester’s home; Lester possessed jewelry from the robbery; Lester himself had made a statement implicating Robert Jones but recanted it before trial; and the crime was part of a series of attacks matching two committed by Lester — and a third similar attack occurred after police had already arrested and jailed Robert. Robert Jones served more than 23 years in jail before a judge overturned his conviction. [Michael Wines / New York Times]

Also in New Orleans, the district attorney’s office agreed to the release of Albert Wolfe after an undisclosed police report cast serious doubt on his murder conviction. In the report, the government’s star witness, and the only witness who tied Wolfe to the murder, gave a vastly different account than he did at trial. At the age of 43, Wolfe was released after spending 22 years behind bars. [Matt Sledge / New Orleans Advocate]

In Orange County, California, law enforcement has run an extensive and systematic snitch operation, secretly placing a few informants in jail next to different defendants awaiting trial, hoping to obtain confessions. The informants, accused of murder themselves, hoped for leniency. The prosecution did not disclose what law enforcement had done. As a result, a judge removed the entire Orange County district attorney’s office from a high-profile death penalty prosecution and precluded the government from seeking the death penalty. In December 2016, federal officials placed the office of Orange County District Attorney Tony Rackauckas under investigation. [R. Scott Moxley / OC Weekly]

In Bexar County, Texas, DA Nico LaHood threatened two defense attorneys after they uncovered potential Brady violations during trial and suggested they would ask for a mistrial and a bar to future prosecution. LaHood claimed he would shut down their practices for requesting such a remedy, and at a hearing, the judge said she feared physical violence. One of the lawyers he threatened defeated LaHood in the 2018 Democratic primary. [Brian Chasnoff / San Antonio Express-News] [Larry Hannan / In Justice Today]

More in Explainers

The Power of Prosecutors

The Power of Prosecutors


We loathe mass incarceration. We loathe police brutality. But most of us have absolutely no idea how to address the critical flaws in our justice system.

What this brilliant short film called “Prosecutorial Accountability” breaks down is the fact that no single individual has more power and more influence in the criminal justice system than the district attorney.

Prosecutorial Accountability

This 4 minute and 20 second animated video, produced by RogueMark Studios of Berkeley and the Alameda County Public Defender's Office, features a narrator discussing the role of prosecutors in the criminal legal system. The video was entirely funded by a grant from the Vital Projects Fund.

Every year in this country, millions upon millions of cases go through their offices. They make sweeping decisions on who to prosecute and who to protect.

“Prosecutors have more power than most us realize,” says the film’s narrator, Bryonn Bain. “Prosecutors alone decide whether or not to charge a crime. They get to decide who is prosecuted — and who gets a pass.”

Sadly, they don’t look or think or feel like the majority of Americans. They are 95 percent white, 81 percent male, and overwhelmingly conservative. We must do better.

More in Podcasts