Hudson County Criminal Lawyer Blog
The New Jersey Supreme Court issued a unanimous decision in State v. Jamel Carlton that has immediate and far-reaching consequences for how extended-term sentences are sought and challenged across the state. At its core, the Court held that a constitutional error occurred when a judge, rather than a jury, made the factual findings required to sentence Carlton as a persistent offender under N.J.S.A. 2C:44-3(a). But the Court did not stop there. It went on to hold that such an error is subject to harmless error review, and that under the specific facts of Carlton’s case, the error was indeed harmless. For criminal defense attorneys and prosecutors working in Hudson, Essex, Union, Passaic, and counties throughout New Jersey, the decision raises important questions about pending cases, ongoing sentencing proceedings, and the long-term viability of the persistent offender statute as it currently reads.
The background of the case is straightforward. Jamel Carlton was convicted after a jury trial of first-degree aggravated sexual assault, burglary, and related offenses. At sentencing, the State moved to have him sentenced as a persistent offender based on two prior New York convictions such as a third-degree robbery from 2007 and a fourth-degree possession of stolen property from 2011. The trial judge, not a jury, made the factual findings establishing Carlton’s persistent offender status and imposed an aggregate forty-two-year term subject to the No Early Release Act. While Carlton’s appeal was pending, the United States Supreme Court decided Erlinger v. United States, 602 U.S. 821 (2024), which held that the Fifth and Sixth Amendments require a unanimous jury, not a judge, to determine whether a defendant’s prior offenses were committed on separate occasions when those findings increase sentencing exposure. Both the State and Carlton agreed that his sentence was unconstitutional under Erlinger.
The constitutional backdrop matters here, and the New Jersey Supreme Court traced it carefully. The line of cases from Apprendi v. New Jersey through Blakely v. Washington and into Erlinger stands for a consistent proposition: any fact that increases the penalty for a crime beyond the prescribed statutory maximum must be submitted to a jury and proven beyond a reasonable doubt. Erlinger extended that logic directly to the persistent-offender context, rejecting the government’s argument that the narrow “prior conviction” exception established in Almendarez-Torres allowed a judge to resolve the surrounding factual questions such as whether the prior offenses were committed on different occasions. The New Jersey Supreme Court made clear that Erlinger abrogated the Court’s own earlier holding in State v. Pierce, which had permitted judges to make exactly those findings.
Where the Court departed from the Appellate Division was on the question of remedy. The Appellate Division had concluded that the complete absence of any jury finding on persistent offender status amounted to something akin to structural error, requiring automatic reversal regardless of the strength of the evidence. The Supreme Court disagreed. Structural errors, the kind that require reversal without any inquiry into prejudice, are reserved for a very limited category of defects, like the total denial of counsel, that render a trial fundamentally unfair at its core. Failing to submit a sentencing factor to a jury, the Court held, is not in that category. Rather, it is the type of error that courts across the federal circuits have consistently subjected to harmless error review in the wake of Erlinger, and the New Jersey Supreme Court aligned itself with that consensus.
Applying the harmless error standard to Carlton’s case, the Court found no difficulty concluding that the error was harmless beyond a reasonable doubt. The predicate facts were entirely undisputed. Carlton’s age, the dates of his prior convictions, the nature of those offenses, and the timeline relative to his current offense were all established by certified documentary evidence that the defense never contested. Defense counsel had in fact conceded all the factual predicates required by the statute. Under those circumstances, the Court reasoned, only one outcome was possible: a jury presented with the same uncontested record would have reached the same conclusion the judge reached. The error was real, but it did not change the result, and the sentence was reinstated.
The more significant long-term takeaway from Carlton is the Court’s pointed message to the Legislature. New Jersey’s persistent offender statute expressly assigns the fact-finding role to “the court,” a deliberate drafting choice made when the statute was enacted to replace the prior Habitual Offender Act. That language is now constitutionally incompatible with Erlinger‘s mandate. The Supreme Court declined to perform the kind of judicial surgery the Appellate Division had attempted and instead called on the Legislature to act promptly. The Court was candid about why urgency is warranted: harmless error review is not a permanent or reliable fix. As time passes and institutional memory of Erlinger and Carlton fades, trial courts, prosecutors, and defense attorneys will continue to encounter persistent offender motions under a statute that gives no constitutionally sound guidance about who makes the findings.
For practitioners on both sides of the aisle, the practical implications are immediate. Defense attorneys should scrutinize every extended-term sentence imposed under N.J.S.A. 2C:44-3(a) where the factual predicates were found by a judge rather than a jury, particularly in cases still on direct appeal or not yet final. Whether Carlton‘s harmless error framework forecloses relief will depend heavily on the record, if the underlying facts were genuinely disputed or the evidence was less than ironclad, the analysis may come out differently. Prosecutors, for their part, should be building the evidentiary record with Erlinger in mind on every persistent offender application and should evaluate whether the facts supporting the enhancement are sufficiently clear-cut to survive harmless error review if challenged.