You've got to hand it to Michael Kinsley; few writers can make a spectacularly flawed argument sound so spotlessly logical. His current critique, in Slate, of the practice of plea-bargaining is apparently derived from a 1978 article in the magazine Public Interest by one John Langbein, entitled, "Torture and Plea Bargaining". To quote Kinsley:

Langbein compared the modern American system of plea bargaining to the system of extracting confessions by torture in medieval Europe. In both cases, the controversial practice arose not because standards of justice were too low, but because they were too high. In medieval Europe, a conviction for murder required either two eyewitnesses or a confession by the perpetrator. This made it almost impossible to punish the crime of murder, which was an intolerable situation. So, torture developed as a way to extract the necessary confessions.

Plea bargaining evolved the same way, Langbein explained. As our official system of justice became larded with more and more protections for the accused, actually going through the process of catching, prosecuting, and convicting a criminal the official way became impossibly burdensome. So, the government offered the accused a deal: You get a lighter sentence if you save us the trouble of a trial. Or, to put it in a more sinister way: You get a heavier sentence if you insist on asserting your constitutional rights to a trial, to confront your accusers, to privacy from searches without probable cause, to avoid incriminating yourself, etc.

This analogy looks reasonable on the surface, but it is in fact wildly self-contradictory. If the modern system of justice is as "larded with....protections for the accused" as the medieval one, then how can plea-bargaining threaten to impose a "heavier sentence" on someone who is nearly guaranteed an acquittal? What is the incentive for the accused not simply to demand a trial, and obtain an exoneration?

In fact, the medieval torture analogy fits one aspect of the modern criminal justice system perfectly--just not the part that involves plea-bargaining. Plea bargains almost always involve cases in which the defendant faces a very high chance of conviction, because the evidence for guilt is overwhelming. In such cases, the defendant has a strong incentive to accept the certainty of a lighter sentence in exchange for the near-certainty of a heavier one. This is a perfectly reasonable transaction for both sides, and not even Kinsley is able to come up with a serious argument against it.

The trouble starts when the authorities have somebody whom they "know" is guilty, but whose prospects for an acquittal are large--either because the evidence is genuinely weak, or because the elaborate labyrinth of "protections for the accused" prevents the damning evidence from being used. In such cases, modern equivalents of medieval justice--from ruthless interrogation techniques to falsification of evidence to "testilying"--become extremely tempting "correctives" to the system. And once used to convict the unmistakably, horribly guilty who would otherwise go free, they become equally tempting tools for winning convictions in cases where the evidence is less clear-cut--including some where the accused later turns out to have been innocent after all.

Now, all of this fits Kinsley's/Langbein's analogy perfectly. It even explains the ostensible topic of Kinsley's column ("Why Innocent People Confess") much better than all of his grumbling about plea bargains. But the clear lesson of this analysis--that doing away with absurd Constitutional barriers to convicting the obviously guilty might actually afford the innocent greater protection, by reducing society's incentives to quietly allow convictions to be obtained by corrupt means--is terribly unappealing to someone of Kinsley's ideological pedigree. So he must instead twist a perfectly good analogy in perverse directions, extracting a lesson--plea-bargaining is bad--that makes no sense, and that benefits nobody except a lot of criminals, their defense attorneys, and their political sympathizers.

Then again, given the incoherence and perniciousness of his claim, Kinsley certainly argues it well.