To suggest that the only reason people take plea bargains is because they don't expect to win in court is hardly reality. Many, many people take plea bargains because of the cost of a protracted legal battle, not to mention the stress.

You're right on one thing though: we have had this discussion before. But if it was a clear-cut as you implicate, there would be no reason Aaron Swartz shouldn't have accepted a plea bargain.

Something that has bothered me about the HN debate on this is that people think that because the MIT security response was weak and easily bypassed, there was no real offense in doing so. This is equivalent to saying that that only those with fortress-like security deserve legal protection from criminal access. Numerous people have objected that because the MIT wiring closet wasn't secure enough then there can't be anything wrong about plugging into it; essentially they're arguing that 'because I could' should be a valid legal defense.

2. Your analogy fails because MIT IT certainly know what they could do to secure their systems more robustly, and chose not to do so. This debate has gone on long enough; most of HN disagrees with your position. Let it drop.

The "Felony" idea comes from the maximum possible sentence for a crime. CFAA has been on the books since like 1986 so as soon as Aaron decided to run afoul of its provisions that was a done deal.

As it turns out many states have provisions to restore voting rights to felons, firearm rights, etc. But this type of case is one of the few things prosecutors don't have discretion on, except by dropping charges completely (which I understand is what you think should happen, but that's not going to match what most of the rest of the country thinks, where fairness of justice and due process are at least theoretically the ideal).

> This debate has gone on long enough; most of HN disagrees with your position.

Oh! Is HN the decider now? I browse this site and I wasn't aware my opinion had been pre-determined for me. Are there any other HN-mandated opinions I should adopt now that this question has been raised? Am I still allowed to like Microsoft products? Are iPhones mandatory or can I keep my Android?

I bet you know how you could secure your home more robustly. That has no bearing on whether someone who breaks in to steal your laptop is committing a crime.

2. You're essentially saying that the victim was asking for it. As for most of HN disagreeing, neither of us knows what the proportions are. There has been some vocal opposition to my arguments, but a lot of it comes from the same small number of people and I've also received a fair degree of support from other quarters. If you don't like it, don't read it.

While I am pro Aaron and anti-MIT this is argument is weak sause.

So what? I also know what I could do to secure my house more robustly (put an electric fence, more cameras, a steel door with secure locks etc), and I don't do it. Does that give you the right to enter into it against my will, much less to a robbery?

>This debate has gone on long enough; most of HN disagrees with your position. Let it drop.

And most of US citizens disagree with you (maybe), and say that abortions should remain illegal. Should you just let it drop? One should not abandon his views because of what a majority (or minority) says. And neither is what the majority agrees with always right.

It isn't designed that way for these offenses, that's the problem. Congress is trying to shove too much under a single offense and then relying on the sentencing guidelines and the courts to separate it out. It shouldn't be possible under a well-drafted law to be guilty of a felony without having done something which is sufficiently egregious to justify a felony penalty.

>You think this allows too much discretion to prosecutors and courts, OK, so did a lot of other people at one point, and that's how we wound up with standardized sentencing guidelines and mandatory minimums.

Both of those things are not what I am suggesting. The trouble is that maximum sentences for nonviolent crimes are too high and the serious offenses are defined too broadly. Establishing or raising minimum sentences does nothing to help that, if anything it makes it worse.

Likewise the sentencing guidelines. The whole concept of "enhancements" is rubbish. All crimes with felony penalties should be narrow, well-defined and require intentional malice. The more broadly defined an offense is, the lower the penalties should be. Enhancements are the opposite: They take minor offenses and ratchet them up into severe crimes based on broad and not necessarily even relevant ancillary factors.

We rarely if ever see people who get into a bar fight be accused of attempted murder and have to rely on the sentencing guidelines to avoid a life sentence, because murder is narrow and well-defined, and things that are slightly similar to murder but not nearly as severe are not covered. They have their own crimes with correspondingly lower penalties. There is no reason we can't do the same thing with computer crimes, drug crimes, etc.

>You don't lose any of your constitutional rights when you accept a plea bargain.

You have a constitutional right to, among other things, a jury trial. You cannot both accept a plea bargain and exercise that right -- you can't take the plea offer contingent on the jury finding you guilty.

>The reality is that people take them because they don't expect to win in court. The standard of proof in criminal trials is very high, so defense lawyers have little incentive to accept a plea bargain unless the evidence overwhelmingly points to guilt.

It makes logical sense to accept a plea bargain in cases where its expected value is better than the expected value of a trial. The issue comes when the outcome of a trial is at all uncertain. If you think you have a 75% chance that the prosecutor will not be able to prove your guilt beyond a reasonable doubt -- or that your interpretation of the law will be adopted by the judge -- but the expected penalty if you lose is five or more times higher than the plea bargain, you are statistically better off to take the plea notwithstanding that you are likely to win in court. Reduce the delta between expected sentences at trial and plea offers and this stops being the case for the accused who have a fair chance at winning while the accused with no chance of winning will still take the plea.

This is kind of tautological. You know that a felony is a crime involving a sentence of more than 1 year, right? It's not reserved for only violent crimes or hardcore criminal offenses, a great many crimes are considered to be felonies. This might seem excessive if you aren't familiar with teh criminal justice system, but it's not that long since we were willing to hang horse thieves and nobody found that idea remarkable.

The trouble is that maximum sentences for nonviolent crimes are too high and are defined too broadly.

Well, how about that CEO in Iowa who robbed $216 million from a pension fund and was sentenced yesterday to 50 years in prison; what do you think an appropriate maximum should be for financial crimes like that? What about the situation where the maximum penalty is not that high, but someone commits multiple offenses?

Likewise the sentencing guidelines. The whole concept of "enhancements" is rubbish. All crimes with felony penalties should be narrow, well-defined and require intentional malice.

Have you ever read the sentencing guidelines or looked at cases where they're argued at appeal? They are very narrowly defined and specific. Go to http://www.sentencing.us/ and look at some examples.

We rarely if ever see people who get into a bar fight being accused of attempted murder and having to rely on the sentencing guidelines to avoid a life sentence, because murder is narrow and well-defined, and things that are slightly similar to murder but not nearly as severe are distinguished and not included. They have their own crimes with correspondingly lower penalties. There is no reason we can't do the same thing with computer crimes, drug crimes, etc.

Has it occurred to you that you might not be very well informed about the law in this area? Because I have read plenty of cases being prosecuted as attempted murder that don't make the news, and likewise both drug laws and the sentencing guidelines are quite a bit more fine-grained than you seem to appreciate.

You have a constitutional right to, among other things, a jury trial. You cannot both accept a plea bargain and exercise that right -- you can't take the plea offer contingent on the jury finding you guilty.

Well, if you don't like the offer, take the trial. The right to a jury is a separate right to that of a trial. Personally I would rather do without the jury part, but that's neither here nor there. The point of a plea bargain is that if you are probably not going to win at trial because the evidence is against you, then by accepting responsibility and saving the state the cost of a trial you are offered a sentence that reflects the reduced burden on the public. It seems to me that you want to have your cake and eat it.

If you think you have a 75% chance that the prosecutor will not be able to prove your guilt beyond a reasonable doubt -- or that your interpretation of the law will be adopted by the judge -- but the expected penalty if you lose is five or more times higher than the plea bargain, you are statistically better off to take the plea notwithstanding that you are likely to win in court.

AT first blush, but when you figure in the opportunity cost of pleading guilty then it's not that good a deal. The vast majority of people who accept felony plea bargains are actually* guilty rather than innocent people who just want to get it out of the way. The rates are different for misdemeanors where a month in jail or a fine of a few thousand dollars may seem relatively convenient. One of the flaws in your analysis is the assumption that the probability of victory is easily calculable. I don't think you've thought this through, because interpretations of the law are usually wrangled over at appeal while trial courts confine themselves to findings of fact and typically rely on model jury instructions for defining the elements of the crimes in question.

* by actually, I mean they have knowingly committed some action that incurs criminal liability, as opposed to inadvertently incurring it due to lack of notice or somesuch.

Reduce the delta between expected sentences at trial and plea offers and this stops being the case for the accused who have a fair chance at winning while the accused with no chance of winning will still take the plea.

But again, the accused don't really know their chances, and if the prosecution or the judge makes an error they may be able to claim a mistrial or an improper conviction, or if the defense screws up in any way they can claim improper assistance of counsel (very popular this one). Basically the guilty have no incentive to skip the trial because they might get lucky. Right now some guy in Illinois who's supposed to be serving a 60 year murder sentence is being sought by law enforcement because he was sent to court to face some additional drug charge, acquitted on a technicality, and mistakenly released from custody [dit: he's been caught]. If you're actually guilty of a crime and all the evidence is stacked against you, you might as well give the trial a whirl: you have nothing much to lose and fighting your case gives you something to occupy your time with, even if the odds of success are not much better than the lottery.

I mean, how exactly do you expect prosecutors to negotiate guilty pleas if they don't actually have anything to offer? That'd be like haggling with someone who just keeps demanding the same price over and over.

It is kind of tautological. What it isn't is consistent with the existing CFAA and the drug laws.

>This might seem excessive if you aren't familiar with teh criminal justice system, but it's not that long since we were willing to hang horse thieves and nobody found that idea remarkable.

The days of hanging horse thieves were also the days when black people were slaves, women couldn't vote and "witches" were burned at the stake, and nobody at the time found those things remarkable either. We've made a lot of progress since then. I believe we can still make some more.

>Well, how about that CEO in Iowa who robbed $216 million from a pension fund and was sentenced yesterday to 50 years in prison; what do you think an appropriate maximum should be for financial crimes like that?

I didn't mean to suggest that all nonviolent crimes should not be felonies, merely that felony nonviolent crimes should be the exception rather than the rule. I don't have a problem with felony penalties for serious financial crimes. Profit motive and intent to cause non-speculative provable high damages tends to remove things from the realm of mistakes that generally honest people can sometimes make.

> What about the situation where the maximum penalty is not that high, but someone commits multiple offenses?

If you mean actually multiple offenses (e.g. someone steals a set of lock picks and then uses them to break into a toolshed; it's two separate offenses), I don't have a problem with that. If you mean someone did one thing but it violates thirteen overlapping laws, that seems ridiculous to me. Prosecutors should not be allowed to charge a single underlying action with more than one crime; pick one and commit to it. Moreover, we shouldn't have such a degree of overlap in the law. It should not be the case that "everyone is guilty of something." That only enables prosecutors to find someone they don't like and then look for (and easily find) something they can be charged with rather than uncovering a criminal act and charging the person who committed it.

>Have you ever read the sentencing guidelines or looked at cases where they're argued at appeal? They are very narrowly defined and specific.

Some of them are narrow and specific. Some of them aren't. The ones that aren't are still problematic. Here are some examples from "Computer Fraud":

A substantial part of a fraudulent scheme was committed from outside the United States? [You don't live in the United States, or your "accomplice" doesn't, or you went on vacation with your laptop a couple of times.] The offense otherwise involved sophisticated means? [She's a witch! Burn her!] Offense involved an intent to obtain personal information? [Almost all computers contain personal information; good luck arguing that it wasn't your "intent" to "obtain" it in any given case.]

Moreover, even for the ones that are individually narrow, the "check the box" cumulative method causes them to become broad in the aggregate. Rather than having a narrow crime and if you didn't do it you go free, we instead get a broad crime that makes everyone guilty and the only question is how many boxes the prosecutor can check to ratchet up the penalty.

>I have read plenty of cases being prosecuted as attempted murder that don't make the news

It doesn't surprise me that prosecutors would file charges like that, but do these non-murderers actually plead guilty to murder or get convicted of it?

>The point of a plea bargain is that if you are probably not going to win at trial because the evidence is against you, then by accepting responsibility and saving the state the cost of a trial you are offered a sentence that reflects the reduced burden on the public. It seems to me that you want to have your cake and eat it.

It makes sense to offer someone a slightly reduced penalty to avoid a trial. When someone is all but guaranteed to be found guilty, the prospect of not having to go through the ordeal (because hey, it's not just the government that feels the burden of a trial) and receiving e.g. a six month sentence instead of the eight months they would get at trial is a fair bargain, and it doesn't significantly coerce anyone inclined to defend themselves into taking the deal. The issue is that the delta is too large: When you get offered six months and the sentencing guidelines say you would get multiple years if you lose at trial on top of your own personal expense of defending yourself, it stops making sense to try to defend yourself unless you are almost certain that you can win, and sometimes even then the burden of the trial (which is multiplied because the stakes are so high it allows you to spare no expense) can still exceed the burden of the sentence offered in the plea bargain.

>AT first blush, but when you figure in the opportunity cost of pleading guilty then it's not that good a deal.

I'm not sure what you mean by opportunity cost. The alternative to a plea is to go to trial and risk a much higher penalty if you lose. You have to discount as best you can the chance of winning by the cost of losing bigger than you would with a plea.

>One of the flaws in your analysis is the assumption that the probability of victory is easily calculable.

Why do you imagine accuracy is necessary for the conclusion? If all you know is that your lawyer tells you that it's more likely than not but also very much not guaranteed that you'll prevail in court, does that leave you in a substantially different decision making posture from knowing that the chances are exactly 75.4%? If anything the inability to predict accurately is likely to make people more risk averse and more likely for an innocent person to accept a plea.

And this calculus is exacerbated by the fact that the prosecutor has about as good an idea as defense counsel how strong their case is and will tailor the plea offer to the strength of their case, with the result that the people more likely to be innocent are still offered plea bargains specifically calculated in order to make them take the offer rather than go to trial.

>Basically the guilty have no incentive to skip the trial because they might get lucky.

I don't understand how you can say that risking an order of magnitude higher penalty is "no incentive" -- the numbers speak for themselves. If there is no incentive then why does almost no one go to trial?

>I mean, how exactly do you expect prosecutors to negotiate guilty pleas if they don't actually have anything to offer? That'd be like haggling with someone who just keeps demanding the same price over and over.

They should be able to offer a small reduction from what the result is expected be at trial. Offering to cut the penalty more than in half is entirely too coercive; the expected penalty at trial should be less than double what plea offers are currently. For the criminals who are caught dead to rights, it would still make perfect sense to take the deal in exchange for a small reduction, because by stipulation they can't win and their lawyers should know that. In contrast, if prosecutors have charged a case where the defendant has a significant chance to prevail, that case should go to court -- and if that happens too often to be economical then either the law needs to be changed or prosecutors ought to stop charging so many questionable cases.

Well no, I think you'll find that those events are rather more spread out over time, for one thing, and that comparing the penalties for specific verifiable acts with systematic but arbitrary discrimination against classes of people doesn't really take your argument anywhere. Stealing a horse is a factual matter which can often be established very easily, and the severe penalty was a rational response to the fact that losing a horse was a often major blow to the victim's livelihood or prospects of survival. This is wholly different to saying someone is inferior because of their ethnicity or gender, or because they have traffic with imaginary beings whose existence cannot be proven.

It should not be the case that "everyone is guilty of something."

But they're not, and people who make the claim that they are tend to cherry-pick their examples. There's a meme that floats around HN to the effect that you could pluck some random person off the street and find enough laws that they're breaking to put them away for years, so we're pretty much living in a police state. This is simply not true.

A substantial part of a fraudulent scheme was committed from outside the United States? [You don't live in the United States, or your "accomplice" doesn't, or you went on vacation with your laptop a couple of times.] The offense otherwise involved sophisticated means? [She's a witch! Burn her!]

Oh FFS. Sophisticated means simply points to the degree of elaboration involved in criminal behavior and thus to degree of intent. It's the same in murder trials; the most severe penalties are given to people who murder by lying-in-wait for a victim or using some method like poison which demonstrates extensive forethought, as opposed to loss of self-control, such as a shooting someone you just had a fight with. These concepts of intent being proportional to criminal culpability are far from new, and they're not vague because they are evaluated in the context of precedent.

Offense involved an intent to obtain personal information? [Almost all computers contain personal information; good luck arguing that it wasn't your "intent" to "obtain" it in any given case.]

It's the prosecution's job to prove intent, and it's very easy to shoot holes in it if it's incidental to the prosecution's main case.

>I have read plenty of cases being prosecuted as attempted murder that don't make the news

It doesn't surprise me that prosecutors would file charges like that, but do these non-murderers actually plead guilty to murder or get convicted of it?

Well, it's kind of hard to get convicted of murder if you are charged with attempted murder, now isn't it? My point is that when you hear of two people being in a bar fight you assume that it wasn't that severe, whereas I find nothing remarkable in the idea that a bar fight can escalate into an attempt on someone's life.

[...] people more likely to be innocent are still offered plea bargains specifically calculated in order to make them take the offer rather than go to trial.

People more likely to be innocent are also more likely to go to trial. Prosecutors don't want to bring cases that they can lose. As with the Aaron Swartz case, and as with virtually every death penalty appeals case, innocense often becomes a proxy for third parties' wishes about how things could be rather than how they actually were.

I don't understand how you can say that risking an order of magnitude higher penalty is "no incentive" -- the numbers speak for themselves. If there is no incentive then why does almost no one go to trial?

You're not following my argument. I'm saying that if we implement your notion of drastically reducing the difference between the offered plea and the sentence a prosecutor will seek at trial, then criminals would have no incentive to avoid the trial. Think about it: if I asked you to plead guilty in return for 10 years, or face the possibility of a ten-year sentence at trial, you'd go trial, wouldn't you?

They should be able to offer a small reduction from what the result is expected be at trial. Offering to cut the penalty more than in half is entirely too coercive; the expected penalty at trial should be less than double what plea offers are currently.

Why a factor of two, based on what? If you are obviously guilty and you don't like the prosecution's offer, you can always reject the plea bargain but plead guilty and make a case in the sentencing phase.

I agree with you that US sentencing in general is too draconian and that prosecutors often abuse their bargaining power. But you're proposing procedural solutions to a problem that is arguably the result of excessive proceduralism, viz. the exhaustive and cumulative nature of sentencing guidelines. You want say you want laws that are very narrowly tailored and specific, and to a large extent that's how things are. We might be better off with more vague laws that left judges more freedom to look at the totality of the circumstances, but some of the time that's going to result in drastic over-sentencing. Procedural specificity is a double-edged sword; if you don't want to see sentences of one day longer than the very specific offenses charges (because that would be unjust) then you have a hard time arguing for sentences that are shorter than the specific offenses indicated by the available evidence.