Plea Bargaining and the Innocent–A response to Judge Kane and Judge Rakoff

In a thoughtful and well written piece the highly regarded John Kane, a long serving federal district judge from Colorado, thinks judges ought to do a lot more to assure that the innocent do not plead guilty to a crime they did not commit.  See John L. Kane, Plea Bargaining and the Innocent, It’s up to judges to restore balance, The Marshall Project (December 26, 2014). This follows another thoughtful piece written by Judge Kane’s friend, the venerable U.S. District Judge Jed Rakoff, from the Southern District of New York, entitled Why Innocent People Plead Guilty.

Both Kane and Rakoff call for a much more active role for federal district judges in assuring that guilty pleas reflect actual guilt rather than a risk calculus by the defendant and his counsel that a guilty plea on conditions set out in a plea agreement is preferable to the risk of trial and a higher sentence.  I have two strenuous objections to this call for greater involvement in the plea process on the part of federal trial judges.

First, both Kane and Rakoff cite the following statistic:

My friend Jed Rakoff writes, “How prevalent is the phenomenon of innocent people pleading guilty? The few criminologists who have thus far investigated the phenomenon estimate that the overall rate for convicted felons as a whole is between 2 percent and 8 percent.” With over 2.2 million people in American prisons that is a haunting amount of injustice.

I don’t know where they get their numbers because they provide no citations, but I do know that those numbers have little do with the federal courts. In short, the judges fail to cite any empirical data that proves there is a significant problem with innocent people pleading guilty in the federal courts.

As both a United States Magistrate Judge and a federal District Judge over the last 28 years I have personally taken more than a 1,000 guilty pleas and I have reviewed many hundreds more taken by Magistrate Judges.  While I do not doubt that a few people plead guilty even though they may not be guilty of the crime charged (although they may well be guilty of a related but uncharged crime), Kane and Rakoff tell us a tale about a bogey man as we figuratively sit around the campfire. But the fearsome creature disappears when the sun rises. Data, gentlemen, where is your data for the federal courts?

Second, Kane and Rakoff in their effort to solve a problem that does not exist in the federal system push for a “mother knows best” role for federal district judges when it comes to plea bargaining. I have always understood our role as trial judges to be a part of an adversarial system that largely depends upon each side making their own decisions about what is best for them. While it is of course true that federal prosecutors almost always have the upper hand, let’s be honest and admit that such dominance arises because the vast majority of federal defendants are caught red-handed. Generally speaking, federal defendants are in a weak negotiation position not because of some flaw in the discovery or plea-taking process in federal court but rather because they are plainly guilty.*

Unless I am misreading them, Kane and Rakoff call for a fundamental change in the role of federal district judges in the plea process. There is no empirical evidence to support a need for this sea change in role. Moreover, their call for a nanny-like approach is in my estimation quite radical. It rejects both the history and tradition of federal trial judges, and the separate and independent roles properly played by prosecutors and defense lawyers. We don’t need, nor should we want, trial judges to become mothers to us all. The federal trials courts are places were big boy pants are worn. That is as it should be.


*This does not mean, of course, that we judges should not enforce Rule 16 aggressively or refuse to call upon prosecutors to adopt an “open file” approach to discovery. Nor does it mean that our Rule 11 plea taking procedures should be lax or performed in a rote manner. Furthermore, we may well want to assure ourselves before trial that the defendant has been fully advised of all plea offers and has knowingly and intelligently rejected them. See, here, for my standing order on that subject in light of Missouri v. Frye.

32 responses

  1. Judge:
    On a semi-related note, how do you feel about Alford pleas, i.e., a plea made for a reason other than the defendant’s guilt?

  2. I’ve disagreed with you more times than I can count but here totally agree with you. Rakoff in particular has an exalted view of the power and wisom imputed to the position he just happens to inhabit, and many of his public pronouncements seem designed to gain approving attention for himself from the NPR set. But mealy mouthed repeating of imaginary facts, even if they fit comfortably with one’s worldview, is a foolish basis for vast policy changes. At least when Congress does this, it’s possible to fix later. And if that doesn’t happen, it may just be that thoughtfull people no less fairminded than the meddlesome judge see it different, in which case the judge should confine himself to applying the law as enacted. Even if that seems beneath him, or unlikely to generate press coverage.

  3. Robert,

    The government opposes Alford pleas in almost all cases. So do I. In the last 20 plus years I think I have accepted only one Alford plea. By the way, I don’t accept no-contest pleas either.

    All the best.


  4. Judge Data where is your data? You offer no data either, if by data you mean statistical studies, which you misdescribe as empirical. Give the paucity of robust studies of almost anything about are legal system, I guess it most be the best in this best of all possible worlds. Incidentally since you invoke your own testimony, because under your usage it is not empirical what is it innate knowledge reawaken in the cave of the law? Nanny judge, please. Do you really believe plea bargaining process is fair bargaining between between parties equal in power, resources, and knowledge rather than a bargain of adhesion. The argument from history and tradition overstates the age of the plea bargain system, unless you assume your judicial career encompass the entire history and tradition of the federal judiciary. The plea bargain system is relatively new and accidental.
    If I was the Freudian you accuse me of being, I might see doubt or even guilt in your language excess. Doubt would not be a bad thing, we act in the face of uncertainty and do the best we can. Guilt no. leave that to the Irish.

  5. Tom it is hard to take seriously someone who does not know the difference between NPR and the New York Review of Books, and it is hardly an exalted view of the judge’s role to ask for a more formal procedure to assure that a procedurally unequal bargain is substantively sound. You might also note that Rakoff seems to suggest changes in FRCrimP, not statutes. The imaginary facts seem to be on your side.

  6. I think it is misleading to view Jude Kane and Judge Rakoff’s proposals outside of the broader problems of pleas bargaining in general. Plea bargaining is out of control. Let’s break down Judge’s Kopf’s reply into two main thesis: (1) That the overwhelming majority of people who plead guilty are in fact guilty and (2) that even if they are not guilty of the specific crime committed they are guilty of some related crime. This position reminds me of the old crack that every wise parents beats their children because even if the parent doesn’t know the crime committed, the child does!

    I grew up in a tradition that justice may grind slowly but it grinds exceedingly fine. Pleading guilty to conduct one did not in fact commit–even if it allows the defendant to escape a harsher punishment–is to put efficiency above truth. It illustrates the lie behind the idea that our judicial system is about truth seeking. If we really cared about truth we would not let someone plead to something untrue just because it was more efficient for the system.

    The second objection I have is the claim that the defendant pleads guilty because he is in fact guilty. This claim is nothing more than judicial vanity. In our system of 100,000 laws combined with the interconnected world we live in everyone is guilty of something, somewhere. In fact, I would go so far to say as the whole notion of innocent until proven guilty is a meaningless concept because there are no legally innocent people anymore–they are only the caught and uncaught, the prosecuted and the unprosecuted.

    I’m not a fan of the specific proposals put forth by the learned judges but let’s remind ourselves who created this problem in the first place–the judicial system. It was SCOTUS who sectioned the whole mechanism of plea bargaining, which was historically an anomalous event up until recently. At least Judge Rakoff and Judge Kane see the problem. Personally I think compulsory prosecution is the better answer–both Germany and Italy have compulsory prosecution, so it’s not as if it can’t be done.

  7. Kenneth. I am an agnostic on plea bargaining, but I think it would be interesting to compare the satisfaction many federal judges take in the plea bargaining system with the many usually from the same ideology who seem endlessly worried about the unfair pressure on corporations to settle class actions, or maybe a comparison of judges worried about plea bargains but not class actions. Most judges, probably are not politicians in black dresses, but joining the cult of the robe clearly is no cure for ideological vision. Compare the 2 letters from former or current federal judges to Rakoff’s proposal. One could doubt they live in the same country.

  8. Any doubts about innocent people pleading guilty in federal court likely flow from the fact that federal crimes are often ambiguous, amorphous and subject to the whims of prosecutors. In other words, a creative AUSA can find a federal crime just about anywhere and failure to do so, rather than indicate innocence, is merely a result of the government not looking hard enough in the right places.

    One only needs to refer to the following to understand the extent of the pathology:

  9. You condemn them for not citing data to support their belief, but then you boldly claim that innocent (meaning innocent of the crime charged) people pleading guilty is a non-existent problem in federal courts. With respect, your ipse dixit is no better.

  10. I am in a poor position to comment here, because I’ve never practiced criminal law. So I’d like to ask a couple of questions: 1) Are there so many guilty pleas and so few jury trials, because federal prosecutors are so expert and dividing the guilty from the innocent, or for some other reason? 2) Relatedly, do you impose stiffer sentences of those convicted after trial than you do under plea bargains? If the answer to the second question is yes, I suggest that you participate in a program that insures that some who are not guilty will be induced to plead, because they have little hope of acquittal (at least sometimes for reasons not related to criminal responsibility), and make the best bargain they can to minimize the time that they will be imprisoned.

  11. Judge, in almost 18 years on the state circuit court bench, I can truly say that I have only one case (out of thousands plead and hundreds tried) where I thought the defendant was “innocent” yet a jury convicted them. I corrected that matter after much soul searching.

    Another issue that you touch on is the term innocent. Of course a jury never makes that determination. I have certainly seen defendants found guilty of murder when I thought they should have been found guilty of manslaughter. But they still committed the “act” which brought them before the court. But only one case (I believe) where the defendant simply did not commit the offense or any lesser included offenses.

    I have a “plea day: where the rule is that a defendant must accept the plea offered by the state or it is withdrawn and the only plea I will accept later (morning of trial) is a “blind plea.” This cuts back on witnesses having to come to court on trial day, fewer jurors being summoned, etc.

    I will accept a “best interest” or Alford plea but that is only where in my judgment the defendant is guilty and for some reason does not want to admit his guilt (a man charged with Robbery 1st does not want to admit dong it when his parents are sitting in the courtroom). In those cases, I have looked at the file and often see:1. a video tape of the robbery with a clone of the defendant holding a gun. 2. a confession. 3. two eyewitness statements identifying the defendant. 4, The return on the search warrant in which the stolen money was found in the defendant’s house. 5 DNA evidence. So while I ask them are they pleading guilty because they think it is in their best interest to do so, I don’t entertain any illusions about them not committing the act.

    Would I take a felony plea from somebody I believed to be innocent, where for example, if found guilty, their habitual offender status required me to sentence them to Life without Parole and the plea offer is “time served with a suspended sentence?” Yes. Juries tend to convict most of the defendant’s we try and if a defendant “knowingly, intelligently, etc” makes a determination that it is not worth the extreme risk to try the case (Life without) as opposed to walking out that day, I think that is a decision that he should ultimately make. Besides, life withouters always bombard me with constant pro se letters and motions so as to make my life miserable. A plea avoids this.

    I often tell defendant, especially when an offer is “time served” and a suspended sentence that I do not want them pleading guilty just to get out of jail but only if they are truly guilty. But then again, I usually have taken a quick look at the state’s evidence and note that it is pretty conclusive.

    As trial judges we don’t try enough cases. I like trying cases which is why I became a trial judge. At the start of my tenure in 1997, I tried a lot more jury trials (probably because lawyers didn’t know where I stood on sentencing and setting aside civil verdicts). My first three years, I think I tried an average of 35 jury trials (civil and criminal) a year. For the year just ending, I believe the number is 15. There are circuits in this state, comprised of 2 or 3 counties, with two circuit judges that try 2-8 jury trials (civil and criminal) total. When I went to the “new judge’s school” there was a section in the book that said the goal of case management was to “dispose of cases, not try them.” Disposing is a term that usually refers to trash. These cases are composed of real people with real disputes and problems. Often a trial provides a catharsis for those involved, win or lose.

    So I think the judges that emphasize low case numbers and disposing of cases are not really doing justice. They are merely bean counters. A now deceased circuit judge who was my best friend on and off the bench (and who I still miss greatly) always pushed to have the lowest case numbers pending. I noticed that when he died that fact was not mentioned in his obituary, or in the eulogy, or anywhere else for that matter.

  12. I don’t see what the fuss is about. In California, we have a rather straightforward and simple procedure for verifying that there is a factual basis for the plea. It does not (in my humble opinion) require too much of the judge or improperly involve the judge in the process. It does not require a great deal of time. It ensures that the defendant has pled to a crime he or she could reasonably have been convicted of by a trier of fact. This post (and the original articles) seem to be trying to make a big deal out of very little.

  13. Tim,

    With respect, I think you are wrong. The data–including the fact that very few 2255 motions (the federal equivalent of a post-conviction attack) are granted– show by most every measure that the huge majority of criminal defendants are guilty. This is evidenced by the defendant’s own recitation of the facts he admit during the guilty plea process, the small number of not guilty verdicts after a jury trial, or the few number of cases where post conviction relief is granted. All the best.


  14. RL,

    If a person desires to change the status quo, the burden of proof normally falls upon the person who wants the change. It is not unfair to ask, “Where’s the beef?” All the best.


  15. Judge,

    I understand why you would prefer that there are more trials. That said, I also agree with you that we almost never see innocent people plead guilty. All the best.


  16. Judge, if you had limited yourself to the burden of proof I think you would have been on stronger ground.

  17. Norm you may be right, but the great bargainer seems a sad substitute for the great mouthpiece, whatever Fallon’s fault, and Perry Mason was named after the first CJ of NE and the make believe Perry never pleaded guilty.

  18. Pingback: The Bestest System Around–Jeff Gamso « Hercules and the umpire.

  19. This is a little off topic but it’s related.

    In popular media it is often depicted that prosecutors offer plea bargain where a substantial part of the consideration is that another party will not be charged/investigated/reduced sentence. Is this really allowed? If so it seems like a horrible injustice in every way imaginable. In particular it undermines the reasoning behind the idea that innocent parties won’t be pressured into plea bargains.

    If it is actually allowed I assume it is quite rare but that is even more reason to forbid the practice since it can’t be really important to the functioning of the legal system. I’m hoping the answer is that like so many other things in the popular media this is just a flat out inaccuracy but I’ve never heard of a particular rule against it.

  20. The problem is that the scales are tipped against the accused. Even with the high burden of proof, the state has every advantage. The state has resources to pay its attorneys. The state has resources to pay investigators (i.e., law enforcement). The state has the threat of throwing the party on the other side into jail. The state is never in jeopardy. The only thing that the accused has is a presumption of innocence, which the prosecutor has discarded because the state believes the accused has committed the crime (as demonstrated by charging a crime). Given these inequities, I would be suprised if the number of innocents convicted was as low as 8 percent…

  21. While I worry that many people who are practically innocent plea out … though primarily in state courts… this is only one reason to worry about plea bargains.

    Plea bargains substantially impair the ability of the court to correct bad law enforcement practices. If the police know that a case is overwhelmingly likely to end in a plea bargain their will be a temptation to use tactics (warrantless search, intimidating witness statements etc..) that would trigger suppression in the court because the case is unlikely to reach the stage where the improper conduct would be discovered but the results of that conduct do serve to convince the defendant to plead out. Even when the person charged is guilty this is a harm.

    Also, I’m afraid that your personal judgement that innocent people aren’t pleading guilty isn’t really that compelling.

    Yes, people for whom there isn’t even a preponderance of the evidence they are guilty will very rarely accept a plea bargain. The people who will accept a plea bargain are those for whom the evidence suggests they are probably guilty…including the case when that evidence isn’t beyond a reasonable doubt (especially if they face automatic triggers that result in large sentences).

    A better question is how many people have plead guilty in front of you where you thought they were guilty but the evidence didn’t seem beyond a reasonable doubt? If this occurs it would seem to be in tension with our judgement that it isn’t appropriate to run the risk of sentencing an innocent person associated with anything less than proof beyond a reasonable doubt. Now one could argue giving people lesser sentences when there is less proof is justified…but it doesn’t seem to be the conclusion we have adopted in our legal system.

  22. Yes, and those very advantages are present in the evidence the judge sees at the time of plea bargain. Indeed, they are accentuated because the state can wait to make it’s case until they have as much evidence of guilt as they can acquire while at the plea bargain stage the defense has had even less opportunity to uncover evidence of innocence than they would at trial.

    Even if the police and prosecutors never get caught up in the game of getting a conviction this creates a statistical bias. If I start investigating someone and stop when I get enough evidence of their guilt to convict them those people who I bring a case against will be likely to include cases where it just happened that I ran across lots of evidence that suggested guilt before I ran across evidence of innocence

  23. Peter,

    Regarding your last point, in a proper Rule 11 inquiry the judge will discuss each and every element of the crime charged, obtain a factual basis for proving each and every element from the prosecutor that typically describes the evidence against the accused, and the presiding judge will ask the defendant whether that evidence is true. In those circumstances, it is hard to believe that there is “reasonable doubt” about factual innocence unless the defendant lies under oath when admitting the underlying facts.

    All the best.


  24. Interesting point. I was defense counsel in the Army and the Army judges did the same as Article III judges for plea bargains.

    you do make a good point that a defendant would have to lie in a plea bargain were he innocent. However, if his choice is between going to trial and facing a hefty sentence should the jury not believe his truthful plea of innocence and a guilty verdict with a short or no sentence, there is a great incentive to lie.

    This brings us to the ethics of plea bargains. If a prosecutor has a weak case for a heinous crime, should the weakness of the case affect the proposed plea bargain? It seems to me that such a position would not be ethical. The crime was as heainous no matter who committed it… but if all the prosecutor wanted was a guilty verdict, he would have gotten it…

  25. Isn’t the likelihood of conviction a valid part of a plea bargain? Indeed, I always thought it was a very important part: the prosecutor is offering a lighter sentence in a case where there is a significant likelihood that a guilty defendant will walk. That presupposes that the defendant did actually commit the crime under circumstances that render him/her legally responsible.

    I do agree that, faced with the possibility–all to real in today’s system–of a very long sentence, people will lie in order to take a plea. I have no idea how often that may occur, but it is very difficult to believe that it does not, or that the frequency approaches the vanishing point.

  26. Judge, those observations would have been helpful in the original post (though I don’t think they settle the matter; if someone is going to plead guilty to avoid an astronomical sentence, even though they didn’t commit the charged crime, it’s not a stretch to believe that they would also go through the hoops of confessing to the charged crime).

    On the other hand, I would not have been able to use “ipse dixit” in my reply, so let’s call it a wash. (Note to fellow lawyers, and judges I suppose: take it from me, it is a BIG mistake to use “ipse dixit” in an argument with your spouse!)

  27. I am very encouraged by the renewed interest of the Supreme Court, and in particular the criticism Chief Justice Roberts, Justice Scalia, and Justice Kennedy, of the DOJ’s policy (in its manual) of charging the most severe offense available under the Code (where, as we saw in Bond, you can charge someone under a chemical weapons treaty/statute for putting household cleaners on a mailbox and car door of a spouse’s paramour), and then using the lengthy sentence and sky-high conviction rate as a cudgel to negotiate down to what was, in all likelihood, a step or two above what the appropriate charge was to begin with.

    It was amazing for the Deputy SG to admit that during an oral argument last term (maybe the grouper fish case), and then to see the justices bring that fact up in one or two subsequent all arguments, to suggest that they will be more inclined to apply a broader rule of lenity in light of that practice. The conservatives on the court are waking up to this problem.

  28. Whether a judge believes that there are (potentially) innocent defendants pleading guilty in his or her courtroom is probably determined by the types of criminal cases that they consider.

    Based on what Judge Kopf has previously blogged about, it sounds like the D. Neb. criminal docket is largely drug trafficking cases, with a fair amount of illegal re-entry cases. Drug possession and illegal re-entry are not likely to involve factually innocent defendants. Illegal re-entry is basically a status crime: we found you here, you’re not supposed to be here, go to jail, be deported. Unless a defendant can prove he or she is a citizen, it’s game over. I doubt there are many U.S. citizens knowingly pleading guilty to illegal re-entry.

    Drug trafficking defendants may have potentially meritorious constitutional defenses (especially Fourth Amendment search and seizure claims), but there are exceedingly few people carrying a kilo of cocaine in the wheel well of their trunk who are factually innocent of a crime. This lady is one of the few exceptions:

    I am not familiar with the dockets heard by Judge Rakoff and Judge Kane. But SDNY is going to be a hub for insider trading charges, where the line between brilliant maneuvering and illegal activity is razor-thin. I can see a defendant charged with insider trading deciding that going to trial was too big of a risk and pleading guilty despite believing they did nothing wrong and their activities staying within the letter of the law.

    I am willing to believe that innocent people are railroaded in state court, where crimes like murder, rape and robbery depend largely on eyewitness testimony and where it can be easy to build a circumstantial case against someone who lacks a solid alibi.

  29. Tim,

    If you don’t charge the crime that is most readily provable, and instead charge a lesser crime, you can wreck havoc with the Guidelines. The Justices know very little about making policy choices for the Executive, and I am not satisfied that they are competent (or authorized) to do so. All the best.


  30. Yes, the stopping rule for investigation does create bias towards charging and convicting the innocent. I am reminded of the caricature of that in the movie, “In the Heat of the Night”, when Chief Gillespie exclaims, “I got the motive which is money and the body which is dead.”

    From the media I get the impression that something of a scientific viewpoint has been embodied in police work, at least in the phrase, “eliminate X as a suspect”. If the police are really trying to find evidence that people are innocent, that is a good thing.

  31. Judge, so you think charging Mrs. Bond with a violation of a chemical weapons treaty/statute for what was a simple assault resulting in an irritated thumb, and charging Mr. Yates under the Enron-inspired anti-shredding law for allegedly tossing back some undersized fish, are examples of the government charging “the most readily provable crime”?

%d bloggers like this: