So between 92% and 98% of the time, plea bargaining gets things right. The "beyond a reasonable doubt" jury trial standard is unlikely to be more than 95% and is in practice probably 90% or lower: http://federaltaxcrimes.blogspot.com/2014/09/proof-beyond-re.... So if the 2-8% error rate is correct, it's not clear that plea bargaining leads to more innocent people being convicted than jury trials.

Moreover, it's unclear to me that you can do much better than 90-95% accuracy. Say you set the threshold at 99.9%, only convicting when the proof is overwhelming. What would be your rate of false negatives (i.e. where you let a guilty person go)? And how would society react to that? There's a false negative rate beyond which people will push back on the court system as being too lenient on the accused. You can already see this in areas of the law where proof is difficult. For example, I think the crazy administrative proceedings for rape cases in college campuses arose, in part, because of the difficulty of holding guilty people accountable in the court system.

There is a huge problem with the US criminal court system (I personally think it's almost entirely product of ludicrous oversentencing, the product of the 1970s-1980s "war on crime"). But I don't think it's helpful to point fingers at parts of the process that are unrelated to those problems.

I think if we dialed sentences back to reasonable levels, a huge fraction of offenders would be able to plead out to non-custodial sentences, and because their BATNAs would be improved, people facing marginal charges would have a much easier time squaring off against prosecutors who would be much more likely to back down if a trial win was still likely to do little more than produce a probation sentence.

It's worth remembering though: most people incarcerated in the US are not incarcerated for drug crimes. Domestic violence and property crimes appear to dominate US metro incarcerations.

That sounds like a classic false dichotomy. How about less plea bargaining and more resources for the courts so they don't have require massive delays to deliver justice.

I doubt you'd feel the current system was just if you experienced what Browder went through. How about a little empathy.

It's perfectly fair that they get a pass for presenting society's side in a trial but when they get to wield plea bargains for the sole benefit of their own career it's nearly privateering.

The real irony here is before I took the case, the Client tried to waive the proceeding motioning the Court for voluntary departure [1], basically agreeing to pay his own way back to his home country without trial, but ICE objected because they wanted to win at trial and have my Client barred from reentry. The twist is, without an attorney my Client won the trial because by law was clear the underlying crime did not subject my Client to removal proceedings...and of course without any legal justification ICE appealed which is around the time I took the case pro bono...after everything my Client still wanted to just voluntarily leave the Country, forego the appeal just to get the hell out of Krome/detention and even having lost the trial ICE and not much of a chance on appeal, against my advice I motioned again for Voluntary Departure and ICE still objected.

[1] http://www.uscis.gov/tools/glossary/voluntary-departure

It would seem to be terribly high, if you take the number of people in jail in the USA as an input and you work out the numbers that comes down to 2% of 2.3M which is 46,000 and 8% of 2.3M which is 184,000!

That should have something in it that allows comparisons to be made.

See: http://www.nytimes.com/2008/04/23/world/americas/23iht-23pri... ("If lists were compiled based on annual admissions to prison per capita, several European countries would outpace the United States.")

What differs is that we keep them there much longer. Cocaine trafficking will get you a few years in Canada in most provinces, versus 5 years minimum in the U.S. and potential 10 or 20 depending on quantity and prior convictions. There is no three strikes law in Canada. In the U.S. many states will give you a life sentence for three violent felonies (in California, even three non-violent felonies).

"Q: How many men and women are in a Canadian prison on a given day? A: In 2010/2011,on a given day 38,219 adults [1] and 16,279 youth (aged 12 to 17 years) [2] were in custody in Canada (federal and provincial), for a total of 54,498 inmates.

2,818 were in custody in British Columbia’s provincial correctional system. [1] 13,758 were in federal custody. [1] In total, just over 24,461 adults were admitted to provincial or territorial jails in 2010/2011. [1]

However there were only 10,916 admissions into provincial or territorial custody to serve a sentence. The difference is accounted for by the number of people held in remand while awaiting their trial or sentencing.[3] "

http://ccphe.ubc.ca/resources/general-information-on-canadia...

That at least answers the question, even though it doesn't agree with what I read earlier.

You also misinterpreted the contents of my link. It's on a tax crime blog, but the text is clear that the author is talking about perceptions of the reasonable doubt standard in general ("In a criminal tax prosecution, just as any other criminal prosecution, the Government bears the burden of proving guilt beyond a reasonable doubt.")

> In a society that values the principles of law and the rule of law, it is better to let a guilty man go than an innocent one, given that they should be innocent until proven guilty.

Every heuristic system of decision will have false positives and false negatives. You can ensure zero false positives, but at the cost of 100% false negatives. If you have any criminal justice system, you'll put innocent people in prison. The social debate is just about how many you're willing to put in prison to reduce the false negative rate to a socially acceptable level.

[1] This exchange in the New York Book Review regarding Judge Rakoff's editorial is also worth reading: http://www.nybooks.com/articles/archives/2014/dec/18/plea-ba.... It contains letters from Robert Swartz (Aaron Swartz's father), and Judge Michael M. Baylson, as well as a response by Judge Rakoff.

I understand that, and what I am saying is that I don't care if the person who wrote the editorial was a judge or any other person, his numbers aren't cited in any way shape or form and are left to rest completely on his laurels. That's fine for an op-ed, but you then latching onto those numbers, even with a qualifier, still undermines your position on the matter.

"You also misinterpreted the contents of my link."

Indeed, I seem to have focused too much on the tax crime angle, so I'll cede this point, but I still think that the numbers backing your linked blog are dubious, eg;

"In one study of his 10 of his colleagues by Judge Weinstein..."

"In a larger survey of federal judges throughout the country... Of the 171 judges who responded"

"ther attempts at empirical studies of these issues show that judges and suggest that jurors are all over the lot on the issue."

All I am saying is this: Show me the empirical studies and not the hugely biased numbers from some small pool of judges being asked how they rate their average conviction probability.

Sorry for the haste and missteps in my original comment.

Your bias comment also makes no sense. The bias in Judge Rakoff's article, to the extent it exists, is to cite higher than a 2-8% number because he's arguing against guilty pleas. And in the surveys the bias runs in the opposite direction: to make the reasonable doubt standard seem more stringent than it actually is. A judge saying that he only requires 75% probability of guilt to sustain a conviction is contrary to his self-interest.

Underestimated?

They're already huge!

1) Police do not take investigating their own lightly and thus most likely have very strong evidence.

2) I don't know the details of these fellows' careers, but presumably they know procedure, know good defense attorneys and would be far more capable of assisting in their own defense than the average citizen.

This is because anytime there is an absolute chance they'll lose, they drop the case.

And if the case is less than a 100% guaranteed slam dunk, they'll ask for a plea so it'll never go to trial.

Prosecutors cherry pick their trials. It's not possible for every prosecutor everywhere to keep a 98%+ average trial conviction rate in a normal system.

Let me rewind a little bit in my own life. I probably could have fought my charges a little more, but the risk outweighed the reward. You go to trial and anything can happen. Was I 100% innocent, no. Was I 100% guilty, no. But because at the time I was arrested my co-defendant had disappeared (I believe he is still wanted but I have never followed up and the past is the past to me) the focus of the prosecutor at that moment in time was squarely on me. The prosecutors offered me a plea that was almost too good to turn down. Looking back, and after talking to a bunch of people, I probably should not have taken the plea, but you can't go through life second guessing your decisions.

Knowing what I know now, IMO, the better your plea deal the less confident the prosecutor feels about the case. I was looking at a minimum 10 year sentence. I took a plea for one year supervised probation. I took the plea for several reasons. First, going to trial would have cost me at bare minimum $50,000-60,000 in attorney costs (contrary to popular belief your ability to obtain a public defender is based on your income, and assets you have access to, and I did not qualify). As it was I had already spent close to $12,000 on attorneys. The second reason I chose the plea over trial was that going to trial would have been a long, drawn out process where the result may or may not have gone in my favor. You are putting your life in the hands of a judge or jury. In the end my decision to take the plea boiled down to cost and knowing that within 1 year I would be done with everything and could move on with my life. If I had gone to trial, it would have cost a lot more and at the time I was finishing probation I would probably just be going to court. I took it as an easy way to lick my wounds, learn from my experience, and put my life back on track quickly.

Funny side story of it all. When I first started discussing costs with my attorney I asked him what his thought was on total costs. He replied that there were two questions he had for me in determining cost. First, did I want him to play golf with both the judge and the prosecutor, or just the prosecutor. And, the second question was how many games of golf I wanted him to play with them. I laughed and told him as many as it takes. Looking back part of me thinks he may not have been joking. Our legal system is full of I'll scratch your back, you scratch mine scenarios.

Reminds me of the time I was learning to drive a stick shift in west texas. My dad was in the passenger seat, and was giving me tons of grief for rolling through a stop sign while paying too much attention to the clutch.

We stopped at a gas station, and a local cop pulled into the pump next to ours. My dad, to tease me, asked the cop how much a ticket for rolling through a stop sign would be.

The cop said "About $100. But let me give you some advice. If you offer the officer the $100 right then and there, you'll save a lot of time and paperwork, and your insurance won't have to find out about it, which can save you more in the long run. It really pays to keep some extra cash in your car."

My dad went poker-faced and said "Thank you for the advice", and then he drove us back home and said "Forget everything that man said.

Unless you want to assume that the court is so broken, it will willy-nilly convict innocents.

But then it doesn't matter much what the prosecutor does, then the justice system is utterly broken.

Racial disparities, emphasis on victimless crime, asset forfeiture, prison overcrowding, and oh, the fact that USA has a much higher incarceration rate than any comparable polity ever in the history of the human race.

E.g. a society is generally entitled to criminalize activities it does not want people to engage in. That some other people may perceive the crimes as "victimless" is neither here nor there. I'd imagine that very few people convicted of drug crimes are not in fact guilty of those crimes, because the evidence in those cases is usually overwhelming (e.g. caught with physical product in hand).

Similarly, the U.S. has had about 5-10x the murder rate of, say, the U.K., going back probably a century or more, long before the recent rise in incarceration rate. So the fact that the U.S. has a high incarceration rate does not by itself mean that innocent people are being put in prison.

And anyone is ever guilty only because we define it to be so. In the state of nature, robbing a store and killing the clerk is simply the stronger animal exercising its free will over the weaker. And if robbery is illegal because we say so, I fail to see how the robber is less guilty of the crime because one country punishes him with 3 years in prison while we punish him with 10.

I don't see "don't smoke or sell pot" anywhere in the Ten Commandments.

For one, they are obligated to investigate any exonerating evidence and circumstances, and any exculpatory circumstances. All of this must be revealed fully to the court and the defense, prior to the main trial.

For another, they must only charge when they believe that a conviction is more probable than an acquittal ("my chances are at best 20%, but let's see how it goes" would be clear misconduct).

In practice, the conviction rates are more than 90%, for sure (if you count a conviction for a lesser crime).

As a prosecutor once told me: trials can be freak events. After listening to the witnesses firsthand, or after some dramatic revelation, everything can look much different than it looked in the file. But other than that, they almost always win at least partially.

And another reason is workload. Apart from really high profile, highly publicized cases, resources are extremely limited and prosecutors mostly try to close the files.

But I think you're right: I think it's desirable that prosecutors let go of most cases in which they're not confident of a win. That's called "prosecutorial discretion". If caseloads are high, we want them working on the clearest-cut violations --- maybe with some public policy stretch cases sprinkled in.

Sometimes that's a risk worth taking, but not often. Acquittals are a bad thing.

What would you do in this situation? Even if he was completely innocent, the temptation to go home right now, to end the nightmare, must be absolutely huge. I can only guess that there are thousands of people like this in US.