Drug Court Oversight of a Key Witness Requires Recusal in Bench Trials Under 1 CMC § 3308(a)

Introduction

In Commonwealth v. Rangamar, 2026 MP 3 (NMI Mar. 16, 2026), the Supreme Court of the Northern Mariana Islands addressed two recurring problems at the intersection of specialty-court practice and criminal adjudication: (1) when a judge’s ongoing Drug Court role with a witness creates an appearance of partiality requiring disqualification, and (2) when due process requires in camera review of confidential Drug Court records sought for impeachment.

Sergio M. Rangamar (Defendant-Appellant) was convicted after a bench trial of assault and battery, assault, and two counts of disturbing the peace, arising from allegations that he fired a pellet gun at Amanda Desebel (the alleged victim and a key witness). While the criminal case was pending, Desebel entered the Commonwealth’s Drug Court Program, and the same judge presided both over Drug Court review hearings for Desebel and over Rangamar’s bench trial.

Rangamar moved (a) to disqualify the trial judge under 1 CMC § 3308(a) due to the judge’s supervisory involvement with Desebel in Drug Court, and (b) to compel disclosure of Desebel’s Drug Court records (or at least to obtain an in camera review). The Superior Court denied both requests, convicted Rangamar, and Rangamar appealed.

Summary of the Opinion

Analysis

1) Precedents Cited and How They Shaped the Decision

A. CNMI recusal framework and “appearance” recusal

• Commonwealth v. Caja, 2001 MP 6 ¶ 2: supplied the abuse-of-discretion review standard for disqualification rulings, framing the appellate posture.
• Bank of Saipan v. Superior Court, 2002 MP 16 ¶ 29 (and the opinion’s later reference to Bank of Saipan v. Superior Court, 2002 MP 17 ¶ 29): anchored the objective “reasonable person, fully informed” test and the restraint principle (avoid “justice shopping” and ensure a judge does not recuse “at the mere sound of controversy”).
• Saipan Lau Lau Dev., Inc. v. Superior Court, 2000 MP 12 ¶ 5: emphasized that Section 3308(a) serves public confidence by requiring recusal whenever circumstances create a reasonable question about impartiality, even absent proof of actual bias.
• In re Abraczinskas, 2023 MP 12: provided the key analytical template—contextual, cumulative-factor evaluation (who was involved, whether the judge interacted with them, whether they would testify, institutional proximity, and promptness/sincerity of the motion). The Court imported that contextual method to the Drug Court setting rather than limiting it to workplace proximity within the judiciary.

B. The boundary of “ordinary judicial exposure”: Liteky and its limits

• Liteky v. United States, 510 U.S. 540, 555 (1994), and In re Estate of Malite, 2011 MP 4 ¶¶ 57-58: established that opinions formed in judicial proceedings “almost never” require recusal absent “pervasive bias.” The Rangamar Court accepted Liteky’s premise but confined it: Liteky protects ordinary impressions derived from adjudication, not exposure in settings lacking adversarial testing.
• United States v. Holland, 519 F.3d 909, 912 (9th Cir. 2008), and United States v. Snyder, 235 F.3d 42, 45 (1st Cir. 2000): supported the “duty to sit” concept that prevents recusal rules from being applied so broadly that routine adjudication becomes impossible. These cases strengthened the Court’s insistence that it was announcing a narrow rule tied to specific, credibility-centered bench-trial circumstances.

C. Objective risk of bias as a due process concern

• Rippo v. Baker, 580 U.S. 285 (2017) (quoting Withrow v. Larkin, 421 U.S. 35, 47 (1975)): supplied the constitutional lens: the question is not “actual bias,” but whether the “probability of actual bias” is too high to be tolerable. Rangamar used Rippo to reject any requirement that the defendant prove subjective prejudice to secure recusal.
• Williams v. Pennsylvania, 579 U.S. 1, 8-9 (2016), and Caperton v. A.T. Massey Coal Co., 556 U.S. 868, 883-84 (2009): illustrated that constitutionally intolerable risk can arise from structural or role-based conflicts, reinforcing the Court’s focus on objective appearance and risk.
• In re Murchison, 349 U.S. 133, 136-37 (1955): underscored the classic principle that adjudicative neutrality must be preserved not only in fact but in appearance—particularly where a judge’s prior role threatens confidence in impartial factfinding.

D. CNMI bench-trial credibility problem: prior exposure to disputed facts

• Commonwealth v. Kaipat, 1996 MP 20: was the Court’s closest CNMI analogue. There, a judge who had learned disputed facts and observed witness demeanor outside the evidentiary record of the later bench trial had to be disqualified because credibility determinations were for the judge alone. Rangamar analogized Drug Court oversight to Kaipat: repeated non-adversarial interactions can create extra-record impressions bearing on credibility in a later bench trial.
• Anderson v. City of Bessemer City, 470 U.S. 564, 575 (1985): was cited for the centrality of credibility determinations, setting up why the bench-trial posture heightened the appearance concerns.

E. Discovery / due process: Brady, in camera review, and confidentiality

• Commonwealth v. Campbell, 4 NMI 11, 15 (1993): identified that Brady encompasses impeachment evidence for significant government witnesses; used here to frame the claim.
• Brady v. Maryland, 373 U.S. 83, 87 (1963), and Kyles v. Whitley, 514 U.S. 419, 437 (1995): supplied the key limitation: Brady reaches evidence within the prosecution team’s control (including police). The Court used this to hold that confidential Drug Court program files are not “prosecution files,” so the dispute was not a Brady suppression problem.
• Commonwealth v. Guerrero: was distinguished rather than applied. Guerrero involved disputed materials in the prosecution’s own files, making in camera review an appropriate mechanism to resolve a Brady-related disagreement. Rangamar held Guerrero does not extend to confidential Drug Court records generated and maintained outside the investigative/prosecutorial process.
• Pennsylvania v. Ritchie, 480 U.S. 39, 58 & 58 n.15 (1987), and United States v. Valenzuela- Bernal, 458 U.S. 858, 867 (1982): established the gatekeeping rule: in camera review of confidential third-party records is not automatic; the defendant must make “some plausible showing” the records contain favorable, material evidence.
• Commonwealth v. Hossain, 2010 MP 21 ¶¶ 13-15, 20: adopted and enforced the same threshold requirement in CNMI law and rejected broad, speculative “might contain impeachment” fishing expeditions. Rangamar treated the request for Drug Court material as Hossain-type conjecture.
• United States v. Bagley, 473 U.S. 667, 682 (1985), United States v. Agurs, 427 U.S. 97, 109-10 (1976), Turner v. United States, 582 U.S. 313, 324-26 (2017), and Kyles v. Whitley, 514 U.S. 419, 434-35 (1995): supplied the materiality limitation. Even favorable information must create a reasonable probability of a different result. Because cross-examination already addressed substance use/credibility themes and no concrete basis suggested a qualitatively different impeachment source (e.g., a prior inconsistent statement about the shooting), the Court found materiality lacking.

2) Legal Reasoning

A. The recusal holding: why Drug Court oversight crossed the line

The Court framed 1 CMC § 3308(a) as an objective, public-confidence safeguard: whether a reasonable person, fully informed, would question impartiality. It emphasized restraint (duty to sit) and Liteky’s “ordinary exposure” safe harbor, but then identified a crucial distinction: Drug Court interactions are (i) frequent and sustained, (ii) non-adversarial and treatment-oriented, and (iii) conducive to forming impressions of honesty, compliance, and reliability without the adversarial testing that structures criminal trials.

Applying the contextual approach of In re Abraczinskas, the Court treated the following as cumulatively dispositive:

• Ongoing supervisory relationship between judge and witness through Drug Court review hearings;
• Frequency and nature of contact (weekly appearances, compliance/progress discussions, monitoring);
• Witness centrality (Desebel was the key witness on identification and events);
• Bench trial posture, where the judge alone assesses credibility.

The Court’s stated narrow rule is the opinion’s core doctrinal contribution:

Importantly, the Court located the case in Section 3308(a) (appearance of impartiality), not Section 3308(b)(1) (personal knowledge of disputed evidentiary facts), signaling that the problem is not necessarily concrete “personal knowledge,” but the reasonable perception that supervisory familiarity and extra-record impressions could influence later credibility determinations.

B. The records holding: why due process did not require in camera review

The Court separated three questions often conflated in practice:

• Is this Brady? No. Drug Court records are confidential program records maintained in a collaborative judicial program (with the prosecution and defense both represented on the “Drug Court Team” under NMI DRUG CT. P. P. § 1(g)(7)), not prosecution-controlled investigative files.
• Even if not Brady, can due process require in camera review? Yes, but only upon a threshold showing under Pennsylvania v. Ritchie and Commonwealth v. Hossain—a non-speculative, plausible basis to believe the protected records contain favorable, material evidence.
• Was that showing made and was materiality shown? No. The request sought broad categories (substance use, “dishonesty,” possible Brady material) without identifying a concrete document, statement, or event in the Drug Court file tied to the charged incident. The Court treated this as impermissible conjecture, and further held that any marginal, cumulative impeachment would not meet Bagley/Turner materiality.

A significant procedural consequence: the affirmed denial of the motion to compel “remains binding on remand,” meaning the retrial/rehearing before a new judge cannot revisit that discovery issue absent some new basis not addressed by the opinion.

3) Impact

A. Specialty courts and criminal adjudication: a clearer recusal trigger

Rangamar is poised to influence how CNMI courts manage dual-role judges in specialty programs (Drug Court in particular, and potentially other treatment/problem-solving courts with ongoing judicial supervision). The holding is fact-bound but operationally important:

• Bench trial sensitivity: The appearance concern is strongest where the same judge is the factfinder and must assess a supervised participant’s credibility. Parties may now more confidently move for recusal when a key witness is currently under the judge’s non-adversarial supervision.
• Narrowness matters: The Court reaffirmed Liteky’s protection for ordinary judicial familiarity. The decision does not imply that any prior contact (or any Drug Court contact) mandates recusal; it targets “sustained” interactions with a “key witness,” in a posture where credibility is central.
• Administrative implications: Courts may need conflict-screening protocols when Drug Court participants become key witnesses in cases assigned to the Drug Court judge, especially for bench trials.

B. Confidential program records: reinforced gatekeeping against fishing expeditions

The decision reinforces that defendants seeking access to confidential therapeutic/supervision records must do more than hypothesize impeachment value:

• No automatic in camera review—the Ritchie/Hossain threshold is a real gatekeeping burden.
• Materiality is independently limiting—even plausible impeachment must be of a kind likely to change outcomes, not merely cumulative.
• Institutional boundary for Brady—Drug Court’s location within the judiciary (and its team structure) matters; the Court treated that structure as cutting against “prosecution possession/control.”

Complex Concepts Simplified

• Bench trial: A trial decided by a judge, not a jury. This amplifies concerns about a judge’s prior exposure to a key witness because the judge alone determines credibility and guilt.
• Recusal / disqualification: A judge stepping aside from a case to preserve impartiality (or the appearance of impartiality).
• 1 CMC § 3308(a) vs. § 3308(b)(1): Section 3308(a) focuses on appearance—would a reasonable observer question impartiality? Section 3308(b)(1) focuses on personal knowledge of disputed evidentiary facts. Rangamar turned on (a), not (b)(1).
• “Reasonable person, fully informed” test: An objective standard that asks how informed members of the public would perceive the judge’s neutrality, not whether the judge feels biased.
• Liteky “safe harbor”: Prior rulings or impressions formed in ordinary judicial proceedings rarely require recusal. Rangamar held that non-adversarial, supervisory settings like Drug Court can fall outside that safe harbor when credibility is later central.
• Brady evidence: Evidence in the prosecution’s possession that is favorable to the defense and material to guilt/punishment, including impeachment of key witnesses.
• In camera review: The judge privately reviews confidential materials to decide whether any portion must be disclosed. It is not automatic; a defendant must make a plausible, non-speculative showing the materials contain favorable and material evidence.
• Materiality (Bagley/Turner): Information is “material” only if there is a reasonable probability the result would have been different had it been disclosed— not merely because it could have helped the defense in some incremental way.

Conclusion

Commonwealth v. Rangamar makes a targeted but consequential clarification to CNMI recusal law: when a judge has sustained, non-adversarial supervisory interactions with a key witness through Drug Court, and the case is tried to the bench with credibility at the center, 1 CMC § 3308(a) requires recusal to protect public confidence in neutrality. The remedy is structural—vacatur and reassignment—even without proof of actual bias.

At the same time, the Court tightened the boundary around confidential specialty-court records: absent prosecution possession/control (no Brady) and absent a concrete, plausible, non-speculative showing of favorable, material evidence (Ritchie/Hossain), due process does not require in camera review—especially where the request amounts to a broad search for possible impeachment and where materiality is not shown.