As many readers of this blog probably know, Aaron Swartz, a young, talented, and prominent figure in the internet world, recently committed suicide after being charged with multiple offenses against the Computer Fraud and Abuse Act and threatened, if he refused a plea deal, with a total penalty of fifty years in prison. I did not know Aaron, although we had a friend in common who has expressed his reaction
to the tragedy in strong and moving terms. Further, I have no definite view on exactly what the relevant law, on I.P. and on access to computers, ought to be, although I have discussed both in writing and teaching.
What I would like to comment on is the issue of overcharging. In theory, most crimes have a range of possible penalties—one to five years in prison, a fine of a thousand to two thousand dollars, or something similar. In practice, prosecutors can and do make a single criminal act the basis for multiple charges—Aaron Swartz was charged with 13 felonies—with at least the possibility of consecutive sentences, making possible sentences far outside the specified range. One reason to do so is to persuade the defendant to plead guilty to one or more of the charges in exchange for the implicit promise of a shorter sentence, presumably what was happening in the Swartz case. This does not seem like a desirable practice—even when used against less sympathetic defendants. I do not know whether what Aaron Swartz did ought to have been punished at all, but I think it would be hard to find anyone, including the prosecutor, willing to argue that it ought to have received the punishment that the prosecutor threatened to impose.
How might one prevent it? One possibility would be to try to eliminate the practice of plea bargaining. On its face, it is a pretty ugly procedure—among other things, a way of getting an innocent defendant to confess in court to a crime he did not commit in order to avoid the risk of a much more severe punishment. In other contexts that would be described as suborning perjury, itself a criminal offense—in this case by the prosecutor.
Eliminating plea bargaining raises two problems. The first is that it would greatly increase the number of trials, since most criminal cases currently are settled out of court. That could be dealt with either by greatly reducing the number of things treated as crimes—eliminating the war on drugs would be a good first step in that direction—or by substantially increasing expenditure on courts. Courts at present absorb only a very small fraction of government expenditure, so the increase would cost considerably less than many things the government now does and could save money by not doing—such as enforcing drug prohibition.
The second problem is how to do it. Plea bargaining involves only an implicit agreement between prosecutor and defendant, which makes it hard to enforce a rule against it. And it makes little sense to refuse to accept a guilty plea.
One possibility, suggested by the previous paragraph, would be to treat offering a plea bargain as subornation of perjury by the prosecutor and punish it accordingly. A more plausible alternative might be a rule under which a defendant could not plead guilty until the prosecutor had entered his charges and the only guilty plea that the court would accept would be to the offenses as charged.
Short of abolishing plea bargaining, how could we make overcharging impossible, or at least not in the interest of prosecutors? One possibility would be to limit prosecutors to treating a single act as a single crime, although defining what was or was not a single act might raise problems. Alternatively, one might permit multiple charges, but specify that if the defendant was convicted of more than one only the most serious conviction would count. But that would not solve the problem of overcharging on a single charge, treating the computer equivalent of a panty raid as interstate transportation of stolen property worth more than five thousand dollars (I am thinking of a famous early computer law case; those who have read The Hacker Crackdown
may recognize it).
My current research on legal systems very different from ours suggests two approaches based on the legal system of Periclean Athens, which I sometimes describe as the legal system of a mad economist—ingenious in ways that sometimes probably worked, and sometimes probably did not. One is to punish the prosecutor for a failed prosecution, at least if it failed badly enough. In the Athenian system, which used very large juries, if the (private) prosecutor failed to get at least a fifth of the jurors to vote for conviction he was fined a thousand drachma. A modern equivalent might be a rule under which a District Attorney who failed to convict on more than a specified fraction of his charges, or repeatedly prosecuted charges that resulted in a unanimous jury vote for acquittal, was automatically removed from office. Like the Athenian rule, it would provide a prosecutor an incentive not to make charges that he did not have a reasonably good chance of proving. Of course, it would also have the disadvantage of giving a prosecutor who had made charges an even stronger incentive than under present rules to convict the defendant, whether or not he was guilty.
An alternative would be the Athenian rule for setting criminal punishment. The prosecutor proposes a punishment, the defendant proposes a punishment, and the jury that has voted for conviction must choose between them. That was, famously, the rule that got Socrates killed. Instead of offering an alternative somewhat milder than execution, such as exile, he first suggested that he deserved a reward rather than a punishment then, at the urging of friends, proposed a fine which his friends were prepared to pay.
Despite the unfortunate outcome in that case, that rule does give the prosecutor an incentive not to seriously overcharge. A jury asked to choose between putting Aaron Swartz in prison for fifty years for what was essentially an act of civil disobedience or giving him a one month suspended sentence and a hundred dollar fine would probably have chosen the latter.