Plea Bargaining Is the Criminal Justice System

The public understanding of criminal justice is built around the trial: evidence tested in open court, a jury, a verdict. That process resolves a very small fraction of cases.

The rest are settled by plea, in negotiations that are largely unrecorded, governed by charging decisions that are unreviewable, and shaped by a sentencing gap between what a defendant is offered and what they face if they insist on the trial the Constitution guarantees.

That gap is the mechanism. When the difference between the offer and the exposure is large enough, going to trial becomes an irrational choice for a defendant regardless of the strength of the case against them, and innocence is not a reliable defense against that arithmetic.

Reform proposals cluster around making the process visible: recording offers, requiring judicial review of the disparity, and publishing charging data by office. None of them reduce prosecutorial discretion, and each would let someone outside the room see how it is being used.

The gap between the public picture and the operating reality is not a secret and has never been one. The caseload tables are published, and the proportion of cases resolved without trial is available to anyone who looks.

What the numbers cannot show is the mechanism, which is where the argument actually lives. A plea is recorded as a defendant's choice, and choice is doing a great deal of work in that sentence.

The structure that produces the choice is the sentencing differential. The criminal code sets exposure by offence, charging decisions determine which offences are in play, and the offer sits somewhere below the exposure.

When that distance is wide enough, going to trial stops being a judgement about the evidence and becomes a wager about the size of the penalty for having asked. A defendant with a strong case and a weak appetite for risk pleads.

This is why the innocence question, which dominates public discussion, is slightly beside the point. A system can produce accurate outcomes and still be one in which the decisive factor is exposure rather than proof.

The procedural framework is not silent. The rules of procedure require a court to satisfy itself that a plea is voluntary and has a factual basis, and judges take that obligation seriously.

But the inquiry happens after the negotiation, on a record the negotiation did not produce. A court can confirm that a defendant understood the offer. It cannot examine how the offer came to be shaped, because that conversation is not recorded anywhere.

Charging is the other unreviewed stage and the more consequential one. The decision about what to charge sets the exposure that makes the offer work, and it is made by an office with wide discretion and no obligation to explain itself.

The published opinions therefore describe a small and unrepresentative slice of the system. Appellate law develops out of trials, and trials are the exception, so the doctrine grows from the cases least like the ordinary one.

The reform proposals are consistent across people who disagree about almost everything else, which is usually a sign they are addressing something real. Record the offers. Require a stated reason where the differential exceeds a threshold. Publish charging data by office.

None of them removes prosecutorial discretion, and none of them should. The discretion exists because cases differ, and a system that charged mechanically would produce worse outcomes than one that does not.

What they would do is create a record where there is currently none, which is the precondition for arguing about any of it. At present the central mechanism of American criminal justice operates almost entirely off the page.