If You Gave Me a Magic Wand (Some Unsolicited Thoughts on Reforming the Sentencing Laws)

The School of Government is a nonpartisan, non-advocacy institution. That’s a tradition I take seriously. If you can find something in the nearly 400 blog posts I’ve written here since 2009 that makes you think otherwise, I hope you’ll let me know.

That said, I am occasionally asked what I would do if someone gave me a magic wand and told me to make our sentencing law better. “Better” can be a tough concept to navigate while staying true to the School’s policy-neutral underpinnings. But I don’t mind sharing a few ideas focused on the mechanics of the sentencing law—largely as a thought experiment designed to call attention to some of the more confusing aspects of existing law.

For all of the suggestions that follow, please assume they would be implemented in a resource-neutral way. In other words, I’m not trying to propose anything that would necessarily increase or decrease North Carolina’s correctional population. Instead, I’m focused on changes that I think would make the law easier to understand and administer.

Suggestion number one: Simplify the felony sentencing grid. After 10 years of teaching, writing, and advising on our state sentencing law, I can tell you that the felony sentencing grid is not . . . intuitive. It’s not rocket science either, but many lawyers and most citizens who look at it for the first time have a hard time translating the numbers they see into a practical understanding of how long a person will be behind bars.

The first surprise for the uninitiated is that the grid-looking part of the grid (the front in the traditional double-sided printing) shows only the ranges of permissible minimum sentences. Until that is explained, a person might look at his Class H, Level I conviction and think, “Ok, I’m looking at 5–6 months.” Nope. Those are just the permissible minimums. The corresponding maximums are on the back of the grid. The sentence (assuming we’re sticking to the presumptive range) will be either 5–15 months or 6–17 months. Showing just the minimums feeds into the perception that the minimum is all anyone will really have to serve in prison—which is largely mistaken for low-level felons.

If I had my magic wand (and by the way, I always wind up Hufflepuff in those online quizzes—sigh), I might suggest getting rid of the back of the grid altogether. How would you do that? One approach would be to reduce the number of permissible sentences down to a relative handful within the existing ranges, and just show those minimum-maximum combinations on the front of the grid. For example, the grid cell for Class H/Level II might show 6–17 as the only permissible minimum-maximum sentence in the presumptive range, 10–21 as the only permissible aggravated sentence, and 4-14 as the lone mitigated option.

There’s an obvious payoff in simplicity, but I realize it comes at the cost of discretion and the ability to tailor the sentence to the individual defendant. If there’s only one permissible presumptive sentence for the defendant, why not just have a computer impose it? The sentence would, to a large extent, be dictated by the prosecutor’s charging decision and whether the State alleged any aggravating factors.

I can think of a few responses to that. First of all, a judge would still have to weigh any aggravating and mitigating factors. Second, the judge would still make the important choice between prison and probation. Third—and perhaps most importantly—the courts don’t take full advantage of their existing discretion in any event. The North Carolina Sentencing and Policy Advisory Commission has long noted that felony sentences tend to fall on one of four “spots” on the range of options available in the defendant’s grid cell: bottom of the presumptive (29%), top of the presumptive (27%), bottom of the mitigated (19%), or top of the aggravated (3%). That means 78% of all terms of imprisonment leave the interstices of each sentencing range unused. The percentage is predictably higher for low level felonies (86% and 87% for Class H and Class I felonies, respectively), where there is often only one “middle” sentence between the spots in each cell. I wrote about the phenomenon here back in 2011.

A single sentencing option in each range would probably be too restrictive for more serious crimes. As you move up the grid and the existing ranges become larger, I could see adding more options within each range. For example, the presumptive options for Class D/Level II could be 59–83, 66–92, or 73–100, giving the court some room to tailor the sentence. Those options could be selected to match historical averages in a resource-neutral way. Regardless of the specifics, the overall objective would be to have the grid present minimum-maximum combinations in a more intuitive way.

Suggestion two: Untangle the post-release supervision component of the maximum sentence from the initial imprisonment portion. Post-release supervision is tricky. There’s no mention of it on the sentencing grid, and defendants don’t generally receive notice of it when they plead guilty. It has become a much bigger issue since Justice Reinvestment, which added post-release supervision for all Class F–I felons in a way that more than doubled the exposure to imprisonment for many of them. (Did you know that PRS revocations now account for more prison entries than probation revocations?)

Defendants and victims often have no idea that most of the stated maximum sentence for a low-level felony won’t be spent behind bars at all, but is rather time built into the maximum sentence to cover the possibility of a revocation of post-release supervision. For example, a person serving a 6–17 month sentence will spend at most 8 months of it in prison. Only defendants who violate post-release supervision will serve any portion of the last 9 months.

When Structured Sentencing was enacted in 1994, it may have been necessary to bake the PRS imprisonment time into the maximum sentence like that. Before a 1995 amendment to the state constitution, the only permissible punishments in the state under Article XI, Section 1 were death, imprisonment, fines, and removal from office. There was a school of thought back then that probation and other suspended terms of incarceration could be imposed only as a by-product of a term of imprisonment, and then only with the defendant’s consent. The 1995 amendment cleared that up, adding suspended prison or jail terms (and restitution, community service, work programs, and other restraints on liberty) to the list of constitutionally permissible punishments.

With that in mind, I think there could be value in bringing post-release supervision out of the shadows and allowing it to stand alone as a probation-like sentence that follows a term of incarceration. If you did that, my proposed one-sided grid could show the imprisonment maximums uninflated by PRS, with the corresponding PRS terms shown separately.

Having the judge impose the post-release supervision term separately, stating both the length of the supervision term and the suspended term of imprisonment that could be activated upon violation, would, I think, be a much clearer statement of what will actually happen in the case. It would allow for a clear statement of the different supervision and imprisonment periods for certain sex offenders—especially the Class F–I sex offenders who have 5 years of supervision but only 9 months hanging over their heads. It might also help defuse a ticking litigation time bomb, discussed here.

A side effect of that approach would be to return the actual imprisonment maximums for low-level felons to where they were before Justice Reinvestment, which would in turn reduce a person’s split-sentence exposure. For instance, a 6–17 month sentence would once again be 6–8, and the corresponding maximum split sentence (measured as one-fourth of the maximum under G.S. 15A-1351(a)) would once again be 2 months instead of 4.25. Depending on how it was structured, the change could also affect how certain shorter felony sentences are viewed as a matter of federal law (an issue discussed here).

Suggestion three: Eliminate the Community-Intermediate distinction. Before Justice Reinvestment, for a sentence to meet the definition of Intermediate punishment, it had to include at least one of six specific conditions of probation (special probation, a residential program, electronic house arrest, intensive supervision, day reporting center, or drug treatment court). A Community punishment, meanwhile, was a sentence to probation (supervised or unsupervised) that did not include one of those six conditions, or a fine only.

After Justice Reinvestment, an Intermediate punishment does not require imposition of any particular conditions. And the only conditions off limits for Community punishment are split sentences and drug treatment court. The bottom line is that there’s less distinction between Community and Intermediate punishment than there was when Structured Sentencing was enacted. In many cases you cannot tell the difference between the two from the probationary sentence itself. In fact, even the computers can’t seem to tell them apart: in FY2016, there were over 3,500 sentences identified as “Community” punishment from grid cells that don’t have a “C” in them. That’s about 20% of all sentences from those cells.

And yet, as discussed here, the distinction matters. It has a bearing on how long probation can be, what delegated authority conditions apply, and whether the four Intermediate probation conditions of G.S. 15A-1343(b4) (including the no-alcohol condition) kick in automatically. Further complicating things, the definitions of Community and Intermediate do not apply in impaired driving cases, which are not sentenced under Structured Sentencing. That’s why, for example, there’s no delegated authority in DWI cases—although pending legislation would change that.

So what if you eliminated the Community-Intermediate distinction? Instead of C/I/A, the dispositional options on the grid could be A/P/F: Active/Probation/Fine Only. We’ve already moved in the direction of an “F” block for some Class 3 misdemeanants. And the notion of a more generalized “P” would jibe with Community Corrections’ movement toward evidence-based supervision based on the results of their risk-needs assessment. That was the rationale behind the 2011 changes to the definitions of Community and Intermediate in the first place. A generalized “P” could also bring Structured Sentencing and DWI probation under the same statutory roof, removing a needless complication from the law. I could envision some statutory tinkering within the “P” to regulate the relative extremes of unsupervised probation and splits.

Suggestion four: Standardize sentence reduction credits. As most readers know, the Division of Adult Correction awards Structured Sentencing inmates Earned Time and Meritorious Time, while impaired drivers get Good Time and sometimes Gain Time. Good Time and Gain Time cut DWI sentences (aside from Aggravated Level One) in half and then some. That is not required by law. It is not because DWI is, as people sometimes say, “sentenced under Fair Sentencing.” It is a discretionary administrative choice of the Department of Public Safety, which sets the credit rules that the jails are obliged to follow.

As Shea noted in yesterday’s post, the substantial credit reductions that apply in DWI cases apparently are factored into judges’ sentencing practices. They factor the credit discount into their sentences, giving 24 months when they want 12, for example. That is confusing to defendants, victims, and other citizens, who probably need to watch a boring bit of YouTube filler to decipher what the sentence really means. It runs contrary to the principle of truth in sentencing that underlies Structured Sentencing’s ironclad minimum.

So the proposal would be to apply to DWI cases the same types of credits that apply in Structured Sentencing cases. Those generally allow for about 4 days of credit per month for misdemeanors. To make the change roughly outcome and resource neutral, of course you’d need to reduce the DWI sentencing rules accordingly. For example, the maximum sentence for a Level One would be reduced from 24 months to 12. It seems like a big change, but it’s not. The difference is that the defendant would actually serve the sentence as announced in court.

I’ll conclude with a couple of quick additional thoughts. One, it might make sense to add an “active punishment exception” for low-level felons, similar to what we have for misdemeanors in G.S. 15A-1340.20(c1). Since 2011, not many felons are truly time served (in that they have served their full maximum sentence). But it happens (I am asked about it almost every week)—and it would happen more often if we adopted suggestion two above and purged PRS time from the maximum. It would be good to have statutory authorization to order an “Active” sentence to time served for defendants with jail credit in excess of the maximum even when there’s no “A” in the defendant’s grid cell.

Finally, I’ll mention one change related to probation, and I’ll keep it brief: I think we could come up with a simpler approach to probation extensions. I’ll save the rest of my probation-related suggestions for another day.

I hope you’ll take these suggestions as what they are: food for thought. I’m not here to lobby for particular changes. But when public officials, policymakers, or advocates want to make changes, I’m happy to help them think of how to do it in a way that makes sense, achieves the desired objective, and avoids unintended consequences.

Surely you have thoughts on these suggestions, or suggestions of your own.


Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.