The Flawed NRC Report: Prosecutors, Plea Bargains, and Long Sentences

(This is Part 9 in my criticism of the recent National Research Council report on incarceration. Part 1 looks at the over-emphasis on the drug war. Part 2 and Part 3 argue against assigning too much importance to longer sentences. And Part 4,  Part 5, Part 6, Part 7, and Part 8 looks at the role of prosecutors.)

As one of my last posts about the NRC report and prosecutors, I want to return to the issue of the relationship between incarceration and longer sentences. In earlier posts, I said that longer sentences have not been a primary engine of prison growth, and that is true if "longer sentences are driving prison growth" is read to mean "people serving more time in prison is driving prison growth." But longer sentences could still be important, just in a way that the report ignores.

They could matter because of how they affect plea bargaining.

Plea bargaining is another one of the report’s strange lacunas: despite being the dominant means by which defendants end up in prison, the term “plea bargain” is used a total of ten times in a 465 page report.* And almost all of these references consider how plea bargaining is used to by prosecutors to evade tough sentencing laws.

But it is clear how tougher sentencing laws can drive up prison populations via plea bargaining, even if time served doesn’t change: as potential on-the-books sentences get longer, prosecutors arrive at plea bargains wielding bigger and bigger hammers.

Arguably one of the Supreme Court’s most important sentencing cases is Bordenkircher v Hayes. Hayes was a minor criminal facing a two-to-ten year charge for check fraud. The prosecutor gave him a choice: take five years, or face life under Kentucky’s then-equivalent (in the 1970s) of a three-strike law if he went to trial. Hayes gambled and lost, and the Supreme Court said that the severity of the “… or else!” did not rise to the level of a due process violation.

The importance of Borderkircher was perhaps cemented by the Court's decision in Ewing v California, which upheld California’s three strikes law against a cruel and unusual challenge. Ewing’s third strike was from stealing three $400 golf clubs, and his relevant predicate felonies were borderline cases. In holding that a 25-to-life sentence for the third strike was not cruel and unusual, the Court indicated that pretty much any non-capital sentence would not face meaningful Eighth Amendment review.

Taken together, Bordenkircher and Ewing provide legislators and prosecutors with tremendous power, power that they have chosen to use in interesting ways. Ewing ensures that legislators are free to pass tough sentencing laws—and thus get that tough-on-crime political boost—without facing much constitutional scrutiny. Bordenkircher means that DAs can threaten to use those laws but don’t have to carry through on the threat—which may actually makes the legislators more willing to pass them. 

After all, if every time the prosecutor threatened someone like Hayes with life he had  to carry through with the threat, state prisons would be overflowing with petty offenders, or prosecutors would never use the tough sentencing laws (if legislatures didn't grudgingly repeal them first). But, as the report itself points out, prosecutors frequently use plea bargaining to evade these tough laws.

Or perhaps many times they don’t so much plead around them as use them to get the sentence they want. In other words, perhaps the prosecutor thinks the “right” sentence for a crime is four years. In a world without Bordenkircher and Ewing, the prosecutor may only be able to threaten, say, six years; with those cases (and the resulting laws passed by the legislature), he can threaten twenty-five. A prosecutor can churn out a lot more four-year plea deals today because of the twenty-five year threat he can credibly invoke. Time served is flat, but the number of inmates rises in no small part because of the unused threat of longer sentences.

And note that the credibility of that threat depends on the moral hazard problem that runs thorugh the criminal justice system. The Bordenkircher threat works only if it is credible: if the counties whose prosecutors make the threat had to pay the costs of carrying through on it when necessary, the threat may not be as credible. But right now, the prosecutor in effect says "take this deal, or I'll spend someone else's money to lock you up for a really long time." That's a credible threat.

Some evidence of this effect can be seen indirectly in William Stuntz's magisterial  Collapse of American Criminal Justice. He points out that in 1974, 17,000 prosecutors handled 300,000 felony cases (about 18 cases per prosecutor); by 2004, 27,000 prosecutors handled about 1,000,000 felony cases (approximately 37 cases per prosecutor). Caseloads more than doubled.** All of which points to prosecutors using bigger hammers to work through bigger caseloads.

The system is one where no-one takes responsibility. Legislators can pass tough laws while relying on prosecutors not to use them. And prosecutors can avoid full responsibility by blaming other county prosecutors for overusing the "penal commons." Moreover, this decentralized blame-shifting is only aggravated by the fact that prosecutors aren't paying for the resources they are using.

Again, the entire goal of realignment in California is to target these latter two problems: to force prosecutors to internalize the costs they are imposing. True, if the legislature passed shorter sentences--the report's preferred policy solution--the problem could be mitigated as well. But the fact that California chose to go a radically different way suggests that perhaps that policy suggestion isn't politically feasible. Or at least that other, more creative alternatives, ones unmentioned in the report, may be more viable.

So perhaps longer sentences do matter. But not in the way that the report thinks. And, as realignment indicates, this difference is more than one of semantics: by understanding better how longer sentences matter, we can craft politically-viable reforms that target the underlying problem.

 

*In the report’s defense, it is not unique in failing to focus extensively on plea bargaining: the Supreme Court tends to discuss only jury-trial rights (Padilla v Kentucky received a lot of attention seemingly because it was a rare decision looking at the plea process), as does most of the academic work in sentencing.

** My guess is that the prosecutor number is the total number of prosecutors, not just those who handle felony cases, so the ratio likely isn't quite right. How this changes the "doubling" statistic, though, is unknowable. If the 10,000-person increase in prosecutors was focused primarily on felony-level prosecutors, then caseloads didn't doube; if focused on lower-level prosecutors, more than doubled.