• SNITCHING: Criminal Informants and the Erosion of American Justice
  • U.S. Attorney General's Guidelines on the FBI's Use of Confidential Human Sources
  • Sarah Stillman, The Throwaways, The New Yorker (2012) (article on the use of juvenile informants)

Recent Blog Posts

Wednesday, August 26, 2009

New ABA opinion on prosecutorial duty to disclose information

The American Bar Association just released an important new opinion regarding the prosecutorial ethical duty to disclose evidence and information favorable to the defense. The rule itself requires prosecutors to "make timely disclosure to the defense of all evidence or information known to the prosecutor that tends to negate the guilt of the accused or mitigates the offense." The opinion interprets this ethical mandate very broadly: it is more demanding than constitutional Brady disclosure requirements; it covers all information favorable to the defense, not just evidence; it is up to the defense, not the prosecution, to evaluate the utility of the information; the government must disclose information as soon as is reasonably practical, and the defendant cannot waive these rights or absolve the prosecutor of her disclosure duties. Here are a few key excerpts:
Rule 3.8(d) is more demanding than the constitutional case law in that it requires the disclosure of evidence or information favorable to the defense without regard to the anticipated impact of the evidence or information on a trial's outcome. . . .The rule requires prosecutors to disclose favorable evidence so that the defense can decide on its utility.
The ethical duty of disclosure is not limited to admissible 'evidence' . . .; it also requires disclosure of favorable 'information'. Though possibly inadmissible itself, favorable information may lead a defendant's lawyer to admissible testimony or other evidence or assist him [sic] in other ways, such as in plea negotiations.
For the disclosure of information to be timely, it must be made early enough that the information can be used effectively. . . . Once known to the prosecutor, [evidence and information] must be disclosed under Rule 3.8(d) as soon as reasonably practical. . . Among the most significant purposes for which disclosure must be made under Rule 3.8(d) is to enable defense counsel to advise the defendant regarding whether to plead guilty.
Where early disclosure, or disclosure of too much information, may undermine an ongoing investigation or jeopardize a witness, as may be the case when an informant’s identity would be revealed, the prosecutor may seek a protective order.
This is an extremely important opinion for informant law and practice, for several reasons...(more after the break)

First is that a great deal of snitch litigation involves so-called Brady or Giglio claims (see this previous post), in which it is alleged that the prosecution withheld information from the defense about snitch witnesses. The ABA opinion not only reiterates the importance of prosecutorial disclosure generally, but explains that the prosecutor's ethical obligation to hand over favorable information is broader than her constitutional obligation. In other words, even if the Constitution does not require disclosure, the prosecutor has an independent professional duty to disclose. While the violation of this duty will not invalidate a conviction, it could lead to professional discipline or firing.

Another interesting feature of the opinion is that prosecutorial supervisors must "establish procedures to ensure that the prosecutor responsible for making disclosure obtains evidence and information that must be disclosed." This includes keeping track of information in one case that might need to be disclosed by a different prosecutor in another case. In other words, prosecutor offices must have data collection and dissemination mechanisms by which their employees can comply with their ethical obligations. This position contrasts with the recent Supreme Court decision in Van de Kamp v. Goldstein this year, in which the Court held that prosecutorial supervisors could not be sued for failing to create data collection systems to provide informant-related impeachment material to defendants. While under Van de Kamp a chief prosecutor cannot be sued for the office’s lack of disclosure procedures, under the ABA opinion she could be disciplined.

In effect, the ABA has decided that the Supreme Court's decisions on prosecutorial disclosure are too weak, ethically speaking, and that prosecutors and their supervisors have far stronger professional obligations to disclose information to defendants, including information about government informants.

Ethical obligations are a crucial feature of the legal profession--attorneys can be disciplined, fired, or disbarred if they violate the ethical rules of their jurisdiction. Disclosure obligations are likewise central to prosecutorial integrity. Attorney General Eric Holder recently threw out the corruption case against Alaska Senator Ted Stevens because Holder concluded that DOJ lawyers violated their obligation to disclose information to the defense. NPR story here. Insofar as states and prosecutors take this new ABA directive seriously, it could be a powerful engine for increased disclosure and transparency.

On a more technical note (non-lawyers may want to tune out here), the opinion does not explicitly address a constitutional issue raised by the Supreme Court in Ruiz. The Ruiz Court distinguished between exculpatory Brady material--material that directly pertains to the defendant’s guilt or innocence--and exculpatory Giglio impeachment material--material suggesting that the state's witness is lying. Classic Giglio includes information about informant rewards, the informant's criminal record, prior history of cooperation or falsehoods, or anything that would impeach the informant's credibility. The Ruiz Court held that although Giglio material is a form of Brady material, the government can withhold that information from defendants prior to the entry of a plea, although not prior to trial. The ABA opinion makes clear, however, that prosecutors cannot wait for trial, but have to disclose information early enough so that defendants can use it meaningfully during plea negotiations.

The question is therefore whether Rule 3.8(d) applies to Giglio impeachment material in the same way that it applies to information that "tends to negate the guilt of the accused." In my view, it does, although I recognize that the ABA opinion does not expressly say yes or no, nor does it distinguish between Brady and Giglio the way the Court did in Ruiz. The opinion does say, however, that a prosecutor's ethical disclosure obligations are broader than her Brady disclosure obligations, which would suggest that the ABA did not think that the Ruiz distinction matters in the ethical context.

Perhaps more fundamentally, the opinion's broad language seems consistent with requiring prosecutorial disclosure of Giglio impeachment. The opinion says that prosecutors must disclose any information favorable to the defense, even if it's not material to the outcome, and that the defense gets to decide on its utility, particularly in figuring out whether to go to trial, plead guilty, or investigate other evidence. These are precisely the sort of decisions that are made based on impeachment material. The ABA even contemplates the situation where the government wants to withhold the identity of an informant: the opinion says that the government can seek a protective order, not that the government can withhold the information. In sum, it would seem anomalous for the opinion to require such broad disclosure, but then permit a prosecutor to withhold the fact that her main witness is being compensated for his testimony and has lied in previous cases.

Monday, August 24, 2009

No Special Treatment for Madoff Cooperator

The Wall Street Journal Law blog posts here that U.S. District Judge Richard Sullivan has refused to let cooperator Frank DiPascali out on bail, even though DiPascali has pled guilty and is helping the government unravel the Madoff scheme. The reason this is newsworthy is that everyone expects courts to treat cooperators well, even when they've committed major crimes (DiPascali's crimes include helping Madoff, lying under oath to SEC investigators, and forging documents--he faces 125 years in prison). In other words, Judge Sullivan is the exception that illustrates the rule. It is more typical for prosecutors and courts to quietly accommodate cooperators--keeping them out on bail, dropping charges, and even helping them with criminal cases in other jurisdictions. In my view, and as I argue in my book, these commonplace accommodations and the culture of cooperation more generally have skewed the criminal system's approach to culpability. Offenders are evaluated as much for their usefulness as their wrongdoing, and even the most heinous crimes have become opportunities for negotiation. For a haunting example, read this story in the Washington Times entitled Drug Dealer Avoids Jail in Daughter's Killing, about a drug informant who avoided punishment for the death of his daughter who died of, among other things, a fractured skull and severe malnurishment.

Wednesday, August 19, 2009

Police Internally Split on Confidentiality Issue

Thanks to Scott Henson from Grits For Breakfast for passing along this important story on a battle raging within the St. Louis police department. Rank-and-file police are refusing to provide information about their snitches to their own police supervisors and city police officials. Here's an excerpt:
Worried about liars in their ranks, city police officials are demanding that up to 20 officers tell bosses details about their confidential informers. But the St. Louis Police Officers Association has won a temporary restraining order to block the inquiry, pending a hearing in court next week. The organization says the probe would jeopardize informers' lives, officers' careers and public safety. At issue is whether officers have attributed fabricated information to confidential informers to obtain search and arrest warrants. Police brass acknowledge in court filings that they believe "one or more" officers "have included false information in affidavits" for warrants, and say the investigation is aimed at stopping "the concerns of police abuse and violation of civil rights."
Ironically, one of the officers' arguments against holding a public hearing is that if informants are called to testify, they will lie. These being the very same informants that police rely on to get the warrants in the first place.

The fact that street cops are at odds with their own police officials on this question reveals some deep dynamics about snitching, including what I call the culture of secrecy surrounding the entire practice. Police and their informants are heavily dependent on one another--police need information while offenders need protection against punishment. Police will often go a long way to protect their sources, famously from defendants and courts, but often from prosecutors and even sometimes from their own police supervisors. This does not mean that police handlers are necessarily corrupt: handling criminal informants inherently means doing unsavory things like ignoring their crimes, bending the rules, sometimes providing addicts with cash for drugs. However, the culture of secrecy makes illegal police conduct that much easier. See this NYT story on Brooklyn police who supplied their informants with drugs. Kudos to the St. Louis police officials who are trying to make the process more accountable and transparent.

Thank You to Other Bloggers

Snitching Blog has gotten some wonderful "welcome to the neighborhood" posts on other blogs. My thanks to, Grits for Breakfast, Hit and Run, Sentencing Law and Policy, and TalkLeft.

Tuesday, August 18, 2009

Committing Crime While Working for the Government

TalkLeft picked up on this story about a Secret Service informant who, while assisting the government, launched one of the largest identity theft operations in U.S. history. Back in 2003, Albert Gonzalez avoided indictment for identity fraud by becoming a snitch; his cooperation resulted in the dismantling of a significant identity theft ring of which he appeared to be the ringleader. He kept on with his criminal activities, however, apparently even using his government connections to warn other hackers.

This is simply one of the biggest problems with informant use: the fact that offenders can use active cooperation not only to avoid punishment but to continue offending. It is a problem inherent in snitching: the most useful informants are typically the most active criminals, so the government has to tolerate some amount of criminality in exchange for information about and access to criminal activities. The scale of the phenomenon ranges: from the small (addicts who stay on the street by providing information to police) to the large (drug dealers who remain in operation by informing on colleagues and competitors) to the mind-boggling (terrorists who provide information to the U.S. government while participating in new terrorist activities). In my book I write extensively about the harm that this practice can cause in high-crime urban communities in particular. When law enforcement tolerates crimes committed by cooperating offenders, whether it is drug use, property crimes, or violence, the neighborhoods in which those offenders live have to put up with it.

Troy Davis Gets a Hearing--Recantation Redux

I posted the other day about how hard it is for defendants to get new trials when the witnesses against them have recanted. Yesterday, the Supreme Court took the unusual step of granting death row inmate Troy Davis a new hearing. Of the nine witnesses who testified that Davis shot and killed Officer Mark MacPhail in 1989, seven have recanted. Of the remaining two witnesses, one--Red Coles--is suspected of being the actual shooter. Here's the NYT story .

While the Davis case is not ostensibly about snitching, it revolves around some classic dynamics associated with informant use and its unreliability. One recanting witness--Kevin McQueen--was in fact a jailhouse snitch who testified that Davis confessed to him. McQueen had worked several times before as a police informant and knew the value of providing information against other inmates. When Queen recanted, he explained that he had fabricated the confession based on jailhouse gossip and television reports. Here's the excerpt from the original appeals brief:
Ex-inmate Kevin McQueen testified at Davis' trial that Troy had confessed to him. McQueen had been a "snitch" in other prosecutions and his version of Troy's "confession" differed wildly from established facts (e.g. Troy was eating breakfast at the Burger King in the morning). McQueen subsequently admitted, "[t]he truth is that Troy never confessed to me or talked to me about the shooting of the police officer. I made up the confession from information I had heard on TV. and from other inmate's talk about the crimes. Troy did not tell me any of this."
This tactic of fabricating other inmates' confessions based on jailhouse talk and publicly available information was made famous by Leslie Vernon White, a Los Angeles jailhouse snitch. In 1989, White went on 60 Minutes and showed reporters how with a few phone calls from the jail he could get enough information to fabricate a confession that police and prosecutors would accept as true. The ensuing Los Angeles Grand Jury Investigation (link to the left) was a response to the White revelations.

Another important aspect of the Davis case that commonly occurs in informant cases is the "first-in-the-door" phenomenon, in which the first suspect to cooperate with police not only gets to direct attention away from himself but can fundamentally shape the official investigation. Red Coles is the man who several witnesses now identify as the real shooter. The day after the shooting, Coles and his attorney went to police and fingered Davis as the shooter--Coles became a witness against Davis at trial. As a result of Coles's cooperation, police resources were directed at Davis. This happens all the time with informants, especially in complex fraud or drug cases--the first suspects to cooperate shape the entire investigation and make it more difficult to discover the truth. Former prosecutor Steven Cohen describes what happens when the government believes a witness who cooperates early:
It is a certainty that the information obtained from the cooperator will become part of the base of information utilized to evaluate future would-be cooperators. Moreover, the information will affect future questioning of witnesses and defendants; it will alter how investigators view the significance of witnesses and particular pieces of evidence; and it may taint the way the case is perceived by the prosecutors and agents. In other words, false information skews the ongoing investigation. The false information may prove critical to issues that have far greater import than whether to accept as true the proffer of another would-be cooperator. Rather, it might impact decisions regarding charges to be filed against other defendants, it might affect decisions related to an appropriate plea for a given defendant, and it might even influence whether the government decides to seek the death penalty. (Steven M. Cohen, "What is True? Perspectives of a Former Prosecutor," 23 Cardozo L. Rev. 817, 825 (2002)).

Monday, August 17, 2009

RSS Feed now available

Here (or click on link to your right).

RSS Feed

Thanks to all the commenters who asked for an RSS feed. Will get it up asap. The blog is still so new we are still working out some glitches.

Friday, August 14, 2009

Informants Killing Informants

To what extent should the government employ and reward murderers, drug dealers, and other criminals as informants? In a developing case in Texas, the U.S. government is trying to figure out who killed one of its Mexican drug cartel informants. Turns out it might have been another U.S.-run informant. Story here.

I bring up this incident because it illustrates a bunch of key issues. One is just a matter of scale: there are now so many informants in the system that we get cases like these in which the government is running the people on both sides of the crime. That's how deep the phenomenon runs.

Second: The government routinely permits serious criminals to remain at large because they are useful, even though they are highly likely to commit new crimes. As one former U.S. special agent remarked about the Texas case, federal officials knew that their informant's job was tracking down people that the cartel wanted to execute. Given that, they "probably should have known he was conspiring to kill someone." Now they're mad because he may have killed one of their other informants. The problem of government-tolerated snitch crime is an old problem. Check out the 2004 congressional report at the left entitled "Everything Secret Degenerates: The FBI's Use of Murderers as Informants." Congress found it appalling that the FBI let known mob murderers remain at large because they were snitching on their rival mafia counterparts. In Chapter Five of my book, I document how the toleration for informant wrongdoing is widespread and can worsen crime and insecurity in inner city communities.

Finally, the Texas story reminded me of Troy Smith. As part of his informant deal, Troy Smith had to produce six arrests of other people in order to avoid drug charges himself. When he tried to sell meth to another informant as part of his quota, he got busted. Because of a procedural mistake by his lawyer, Smith could not raise the "public authority" defense, i.e. the claim that the government authorized him to commit the crime. Smith is currently serving a 12-year sentence, arguably for doing exactly what the government told him to do. I tell this story not only because it seems ironic and unfair, but because the pervasive use of informants invites precisely this kind of debacle.

Thursday, August 13, 2009

In the News--Recantation

When a criminal informant recants his testimony after a defendant has already been convicted, it is typically very difficult for that defendant to get a new trial. This happens more often than you might think--informants change their stories all the time, but the rules of criminal procedure and habeas corpus make it very hard to upset the original conviction. Today's New York Times reports on Fernando Bermudez, a man who tried 11 times to get his 1992 murder conviction overturned after the main witness recanted. A new judge has finally held that he might be entitled to a new trial. Mr. Bermudez also has the good fortune to be represented by my exceptionally skilled former colleague Barry Pollack, partner at Miller Chevalier.

Wednesday, August 12, 2009

Pleading Guilty versus Going to Trial

Sometimes I will post about an aspect of what I call informant law - i.e. the legal rules and policies that govern the use of informants. An important but little-known fact about the rules of snitching: defendants who go to trial are constitutionally entitled to negative information about informants who might testify against them (usually referred to as impeachment or Giglio material), while defendants who plead guilty (approximately 95 percent of all felony defendants) are not entitled to this information.

Lets say Defendant Jane Smith is accused of dealing drugs, based on the say-so of criminal informant John Doe. If Smith goes to trial, the government is obligated to give her any material information in its possession regarding Doe’s credibility, including the rewards he got for cooperating against Smith, his prior convictions, instances of perjury or recantations, and things like that. The Supreme Court has held that this is necessary to ensure a fair process. But the Supreme Court also held in United States v. Ruiz that if Smith takes a plea, she has no right to see that information. This means she has to decide whether to plead guilty without knowing how credible or corrupt John Doe might be. She only gets to learn that information if she rolls the dice and goes to trial.

Ruiz is about what the U.S. Constitution requires--other rules may come into play. For example...

some states and districts demand more disclosure, and require the government to provide impeachment material to defendants to be considered during plea negotiations. Some federal prosecutorial offices provide informant impeachment material voluntarily as a matter of internal policy. But the bottom line is that, constitutionally speaking, they do not have to.

This rule has some significant consequences. One is that the government can insulate shady informants by offering defendant good deals. Thats what happened in Ruiz - the government offered Angela Ruiz a so-called fast track plea if she would waive her right to impeachment material about the snitch in her case. If a defendant is sufficiently scared of going to trial-maybe her lawyer doesn't have time or resources, or maybe she has a prior record and can't testify-she may take the deal to avoid a worse sentence.

More broadly, it means that the government can expect that in the vast majority of cases it will never have to disclose the deals it makes with its informants, or the kinds of people that it uses as informants, because over 90 percent of cases are resolved by plea. When defendants don't get to see this material, the public doesn't either.

This is a problematic way to run a criminal system that is ostensibly committed to transparency and public adjudications of guilt. When informant deals stay secret, the public loses sight of how police and prosecutors evaluate crime and impose punishment. Many criminal informants escape liability for very serious crimes - Ruiz makes it easier for the government to hide this fact. When information sources are shielded from scrutiny, moreover, we don't get to see how the government investigates crime or chooses its targets. These are important aspects of criminal justice, but the nature of snitching rules like the Ruiz decision tends to erase them from public view.

Tuesday, August 11, 2009

Professional Prison Snitch Ring

I recommend this recent feature article in Reason Magazine by Radley Balko, entitled Guilty Before Proven Innocent. It tells the mind-blowing story of an innocent family in Louisiana, Ann Colomb and her three sons, who were wrongfully convicted of drug trafficking based on the testimony of numerous prison snitches. The informants were part of an information-selling network inside the federal prison, in which inmates purchased files and photographs to help them fabricate testimony which they then marketed to prosecutors in order to get sentence reductions. A bunch of inmates got hold of the Colomb file, and told prosecutors that they would testify against the family. If it werent for a few chance encounters that revealed the scam, the Colomb family would still be in federal prison.

I like this story because it highlights some classic problems with criminal informants. It also illustrates the scale of the phenomenon--and its potential for massive miscarriages of justice-- in ways that may be surprising to people unfamiliar with the daily workings of the criminal process.

As the story illustrates, criminal informants are a primary (and infamous) source of wrongful convictions. Check out the link to the Northwestern University Law School report entitled The Snitch System on the left. Second, there are a lot of them: the government planned to use dozens of prison snitches against the Colomb family, and presiding Judge Tucker Melancon indicated that the phenomenon was pervasive. Third, prosecutors rely heavily on them even when the government should be suspicious. The prosecutor in the Colomb case did not appear to know that his prison snitch witnesses were selling information to each other and then lying about it; rather, he took them at their word even though he knew they had massive incentives to lie. Perhaps most importantly, the story shows how snitching has become commonly understood as a way for suspects and inmates to game the system. The Louisiana snitch ring sold information for thousands of dollars inside and outside the prison. This business plan was a response to a central fact about the U.S. criminal process--that information and leniency are traded freely between offenders and the government without rigorous fact-checking. This case just took it to a new level.

Welcome to Snitching Blog

snitching = when police or prosecutors offer lenience to criminal suspects in exchange for information or cooperation
Snitching Blog is about a part of our criminal system that most people know little or nothing about: criminal informants, or snitches. At any given moment, thousands of informants are trying to work off their own criminal liability by giving information to the government. These informants may be in court, in prison, on the street, or in the workplace. Police and prosecutors often rely heavily on information obtained from snitches--especially in drug enforcement but also in white collar crime, organized crime, and terrorism investigations. In fact, it is impossible to fully understand the U.S. legal system without understanding snitching. Nevertheless, there is very little public information available about this important public policy. That's where Snitching Blog comes in.

This blog does a bunch of things. It discusses how snitching works--on paper and in reality. It provides resources to individuals, lawyers, law enforcment, and legislators--check out the links on the left. It covers current events and news stories. And it lets you share your own experiences by posting a "Testimonial"--click on the link at the right.

Please look around. Suggestions are welcome.