State v. Kim: Supervisory Limits on “Consciousness of Guilt” Jury Instructions (and a Call to Abolish Them)

1. Introduction

State v. Kim (Conn. July 28, 2026) reached the Supreme Court of Connecticut on the defendant’s challenge to a trial court’s “consciousness of guilt” jury instruction—an instruction that invites jurors to consider whether certain post-event conduct (such as inconsistent statements) may reflect guilt. The defendant, Yong Sik Kim, sought (1) reversal of his conviction because the instruction was erroneous and harmful and (2) a prospective rule eliminating consciousness-of-guilt instructions altogether.

Justice D’Auria (joined by Justice McDonald) filed a concurrence and dissent. He agreed that the instruction given at Kim’s trial was an abuse of discretion, but he parted ways with the court’s approach in two key respects: (a) he rejected any suggestion that supervisory-power “limits” or stare decisis meaningfully constrained the court from abolishing the instruction; and (b) he disagreed that the error was harmless, concluding he lacked “fair assurance” that the instruction did not affect the verdict.

The opinion is also significant for its description of the majority’s new supervisory framework: trial courts should give a consciousness-of-guilt instruction only when a tailored charge will assist the jury beyond argument and general instructions, the benefits outweigh the risks, and—critically—either the state can explain why the instruction is necessary or the defense requests it/does not object. D’Auria viewed that reform as highlighting why the instruction should be discarded entirely (or, at minimum, limited to defense-requested or unopposed situations with a balanced charge).

2. Summary of the Opinion (Concurrence & Dissent)

  • Agreement with the majority on error: The trial court abused its discretion by giving a consciousness-of-guilt instruction that overstated weak/ambiguous evidence and adopted the state’s “pejorative” framing that the defendant “change[d] his story.”
  • Disagreement on supervisory authority and stare decisis: D’Auria argued nothing about supervisory power doctrine or stare decisis should stop the court from eliminating consciousness-of-guilt instructions, just as the majority felt free to sharply curtail them.
  • Preferred rule: Abolish the instruction altogether; alternatively, restrict it to cases where the defense requests it or does not object, and require a neutral, balanced instruction that fairly states both parties’ characterizations.
  • Disagreement on harm: D’Auria concluded the case was close and the instruction likely mattered because the case turned on credibility, the inference was weak, and the jury returned a split verdict (acquittals on four more serious counts), suggesting the jury did not fully credit the complaining witness and could have been improperly influenced on the counts it did convict.

3. Analysis

3.1. Precedents Cited

A. Supervisory authority: prospective rules vs. reversal as a remedy

The concurrence’s doctrinal center of gravity is the distinction—drawn most explicitly in State v. Carrion—between supervisory authority used prospectively to guide trial courts and supervisory authority used retrospectively to justify the extraordinary remedy of reversing an otherwise valid conviction.

  • State v. Carrion: D’Auria relies on Carrion’s two-category taxonomy. Category one permits the court to “articulate a procedural rule as a matter of policy” without reversal; category two uses supervisory authority to “remedy a perceived injustice” and requires a showing that reversal is “truly necessary” to justify an “extraordinary remedy.” D’Auria’s point is that, once the majority found the error harmless, the remaining question (what rule should govern future cases) belongs in category one—where “prudence and good sense” is the relevant standard.
  • State v. Elson and State v. Diaz: Cited to show the court’s recent efforts to clarify the categories and standards, reinforcing that prospective supervisory guidance is not governed by the same “extraordinary remedy” threshold.

B. Supervisory authority frequently operates through jury instructions and canvasses

D’Auria situates the consciousness-of-guilt debate in the court’s broader practice: Connecticut often uses supervisory authority to refine instructions that risk misleading jurors, even absent reversal in the case at hand.

  • State v. Medrano: Discontinued an instruction permitting jurors to consider the defendant’s interest in the outcome when testifying, illustrating willingness to remove long-used charges due to unfair prejudice and limited utility.
  • State v. Ledbetter (overruled in part on other grounds by State v. Harris): Required an instruction addressing the risks of misidentification in eyewitness evidence, highlighting the court’s proactive role in guarding against common jury error.
  • State v. O'Neil: Revised and directed the future use of a specific “Chip Smith” charge.
  • State v. Aponte, State v. Griffin, State v. Devalle, State v. Schiappa: Examples where the court prohibited future use of specific formulations (e.g., “two-inference” charge; reasonable doubt definitions; intent presumptions) because they could mislead or dilute constitutional protections.
  • State v. Connor, State v. Gore, Duperry v. Solnit: Canvass cases showing supervisory authority also shapes procedure beyond jury instructions.
  • In re Daniel N.: Cited for the proposition that supervisory authority is reserved for serious matters implicating systemic integrity and fairness.

C. Stare decisis: flexibility for procedural/evidentiary rules and jury instructions

D’Auria argues stare decisis does not meaningfully constrain abolition of the instruction—especially where the majority itself substantially departs from prior practice.

  • State v. Salamon: Stare decisis is not “an insurmountable barrier”; consistency must serve “a justice related end.”
  • Payne v. Tennessee: The concurrence invokes the principle that stare decisis is weakest for procedural and evidentiary rules—precisely the terrain of jury instructions.
  • State v. Malave: A key analogy. Malave abandoned the missing-witness rule and instruction in criminal cases even while “mindful” of stare decisis—supporting D’Auria’s claim that long usage does not immunize an instruction from abolition.

D. Prior Connecticut consciousness-of-guilt cases (as evidence of how much the majority changed)

D’Auria contends the majority’s new approach effectively displaces numerous prior decisions upholding instructions given over defense objection, in state-favoring form, or with pejorative language—undercutting the claim that stare decisis should prevent outright abolition.

  • State v. Cooper
  • State v. Coward
  • State v. Luster
  • State v. Figueroa
  • State v. Hines

E. Out-of-state skepticism of flight/consciousness-of-guilt instructions

To support abolition, D’Auria canvasses decisions warning that such instructions overemphasize a single inference and improperly comment on evidence.

  • United States v. Mundy (Second Circuit): Characterizes flight instructions as a disfavored vestige of judge-comment practice.
  • United States v. Robinson (D.C. Circuit): Suggests the issue is best left to counsel’s argument with minimal judicial comment.
  • Rob- bins v. People (Colorado): Flight instructions are “rarely advisable.”
  • Fenelon v. State (Florida): Comment on flight should be reserved for counsel.
  • Renner v. State (Georgia): Holds it error in future cases to charge on flight.
  • Dill v. State (Indiana): Flight instruction error due to potential to mislead.
  • State v. Marsh (Iowa): Ordinarily should not draw attention to specific evidence.
  • State v. Cathey (Kansas): Declares it “clearly erroneous” to instruct on consciousness of guilt by flight, concealment, fabrication, or false information.
  • State v. Grant (South Carolina): Flight charge places undue emphasis and should not be used.
  • State v. Menard (South Dakota): Instruction should be used “sparingly.”
  • Hawkins v. State (Texas): Flight instruction improper as a comment on weight of evidence.
  • State v. Jefferson (Washington): Discards instructions emphasizing specific circumstantial evidence.
  • Hadden v. State (Wyoming): Announces giving a flight instruction in future cases will be reversible error.

F. Harmless error and the meaning of “fair assurance”

D’Auria frames the harm inquiry through Connecticut’s standard for nonconstitutional instructional error.

  • State v. Edwards: Nonconstitutional instructional errors are harmless if not “reasonably probable that the jury was misled.”
  • State v. Adam P. and State v. Boyd: The defendant bears the burden, but the appellate court must have “fair assurance” that the impropriety did not substantially affect the verdict.
  • On the “split verdict” inference, D’Auria relies on cases cautioning against assuming acquittals prove harmlessness: State v. Iban C. and State v. Maguire.
  • State v. Ancona: Recognizes the general presumption that jurors follow instructions, but D’Auria argues that presumption does not answer whether an erroneous instruction mattered.

3.2. Legal Reasoning

A. Why the instruction was erroneous (as applied)

The trial court’s instruction told jurors the defendant “initially” denied being at the complainant’s home, “only to change his story” by later admitting he had been there. D’Auria agrees with the majority that this was improper because:

  • The underlying “consciousness of guilt” evidence was attenuated: it arose from inconsistent statements in an employer’s internal inquiry, not a clear false denial to police; and testimony left uncertainty whether a true inconsistency even occurred.
  • The instruction’s “change his story” phrasing carried a pejorative connotation and risked signaling that the court endorsed the state’s inference of deception and guilt, rather than leaving competing inferences to jurors.
  • The instruction marshalled evidence in a one-sided way—highlighting only the state’s characterization, despite the defendant’s alternative explanation (a promise to the complainant not to tell coworkers).

B. Supervisory authority: D’Auria’s critique of the majority’s constraints analysis

D’Auria’s core institutional claim is that the majority, having found harmlessness, should treat the prospective rulemaking question as a straightforward category-one supervisory exercise under State v. Carrion—guided by “prudence and good sense.” In his view, it is mistaken (or at least unnecessary) to invoke the demanding standards associated with reversing convictions when the court is simply deciding what future trial courts should do.

He then turns that institutional framing against the majority’s compromise: if the court can (and does) sharply narrow when such instructions may be given, it can also abolish them. The concurrence portrays the majority’s new test as both (1) a substantial break from precedent and (2) evidence that the instruction is expendable, duplicative of general circumstantial-evidence/credibility instructions, and unfairly prejudicial.

C. Stare decisis: why abolition would be no more disruptive than “near-elimination”

D’Auria argues stare decisis does not justify preserving even “fragments” of an instruction that the court now deems risky. He emphasizes:

  • The court has repeatedly modified or abolished longstanding instructions under supervisory authority (e.g., State v. Medrano, State v. Malave) when fairness concerns outweigh continuity.
  • The majority’s own new criteria (requiring state justification upon defense objection; cautioning against labels and certain wording; recognizing past cases’ lack of clarity) necessarily displace much prior practice, making it inconsistent to invoke stare decisis as a reason not to go further.

D. The dissent’s preferred rule: abolition (or defense-controlled use)

D’Auria’s policy reasoning is candid and pragmatic. He stresses that a “neutral judge” highlighting a guilt inference can cause jurors to give that inference undue weight—especially because it “strikes closest to the ultimate issue” (guilt). He favors leaving the issue to counsel’s argument, consistent with United States v. Robinson and the broader judicial disfavor discussed in United States v. Mundy.

In the alternative, he offers a narrower retention model: allow the instruction only when the defense requests it or does not object, and require a balanced recitation of both sides’ characterization—thereby avoiding judicial endorsement of a prosecution inference.

E. Harm: why D’Auria could not find “fair assurance”

The concurrence’s harm analysis ties the instructional error to the trial’s decisive feature: credibility. D’Auria highlights that the jury acquitted the defendant on four more serious counts (including first-degree sexual assault and second-degree strangulation), indicating the jury may not have credited the complainant’s account in full. In that posture, an instruction that effectively branded the defendant as having “change[d] his story” could have tipped the balance on the counts of conviction.

D’Auria rejects the notion that the model instruction’s standard caveats (“no presumption”; jurors “may, but are not required to” infer guilt) cure the problem when the court simultaneously (and improperly) endorses the state’s narrative framing. He also argues that general instructions on circumstantial evidence and credibility do not mitigate the prejudice; in context, they may underscore the court’s special emphasis on the consciousness-of-guilt inference by giving it a separate spotlight.

3.3. Impact

A. Doctrinal impact (as described in the opinion)

Although D’Auria would have gone further, his writing spotlights what the majority’s new framework effectively does: it transforms consciousness-of-guilt instructions from a relatively routine discretionary charge into a tightly constrained tool—one that must be justified as genuinely helpful beyond argument and general instructions, and that carries heightened sensitivity to one-sided phrasing and judicial endorsement.

B. Trial-level practice

  • Fewer such instructions: Courts are incentivized to omit them unless a clear, case-specific need can be articulated.
  • More litigation over “necessity”: The requirement that the state explain why jurors need this instruction (especially over objection) invites motions, offers of proof, and appellate records focused on the instruction’s marginal utility.
  • Drafting discipline: The case underscores that pejorative phrasing (“change his story,” “guilty conscience”) and one-sided summaries create reversible-risk conditions, particularly in credibility-driven cases.

C. Appellate review and harmless error

D’Auria’s harm discussion signals a warning for future harmless-error analysis: where the case turns on credibility and the instruction adopts the prosecution’s characterization of ambiguous conduct, appellate courts may have difficulty concluding with “fair assurance” that it did not matter. His reliance on State v. Iban C. and State v. Maguire suggests skepticism toward treating split verdicts as a proxy for harmlessness.

4. Complex Concepts Simplified

  • “Consciousness of guilt” instruction: A jury instruction telling jurors they may (but need not) infer guilt from certain conduct after an alleged crime (e.g., flight, concealment, inconsistent statements). The controversy is that it can overemphasize one inference and sound like the judge is endorsing guilt.
  • Supervisory authority: The Connecticut Supreme Court’s power to manage and improve trial procedures to ensure fairness and integrity. Under State v. Carrion, the court distinguishes between using this power to set prospective rules (no reversal needed) and using it to reverse a conviction (an “extraordinary remedy” requiring a showing that reversal is “truly necessary”).
  • Stare decisis: The principle of adhering to precedent for stability. D’Auria emphasizes it is weaker for procedural/evidentiary rules (citing Payne v. Tennessee) and does not prevent abandoning longstanding instructions when fairness so requires (e.g., State v. Malave).
  • Marshalling the evidence: When a judge summarizes or highlights specific evidence. Modern practice disfavors it because jurors may treat judicial emphasis as guidance on what matters most.
  • Harmless error / “fair assurance”: Even if an instruction was wrong, a conviction stands if the reviewing court has “fair assurance” the mistake did not substantially affect the verdict (drawing on State v. Boyd; and the “reasonably probable the jury was misled” formulation in State v. Edwards).

5. Conclusion

Justice D’Auria’s concurrence and dissent in State v. Kim treats the case as a turning point in Connecticut’s approach to consciousness-of-guilt jury instructions. He agrees the instruction here was an abuse of discretion because it overplayed equivocal evidence and implicitly endorsed the state’s narrative. He would go further than the majority by abolishing the instruction (or at minimum cabining it to defense-requested/unopposed situations with a balanced charge), arguing that supervisory authority and stare decisis pose no real barrier to doing so and that the instruction’s risks routinely outweigh its marginal benefits. On the case-specific outcome, he underscores that in a credibility-driven trial with a split verdict, the court should not be confident the erroneous, one-sided instruction was harmless.