delivered the Opinion of the Court.
T1 Rhoderick Flockhart was convicted of distribution of marijuana and possession of eight ounces or more of marijuana. On appeal, the court of appeals ruled on three *230issues. First, a divided court of appeals reversed Flockhart's convictions because the trial court instructed the jury that it could discuss the case before deliberations in violation of his constitutional rights. See People v. Flockhart, -- P.3d --, --, --, No. 07CAS12, slip op. at 9, 19 (Colo.App. Dec. 24, 2009) (selected for official publication). In dissent, Judge Bernard would affirm Flock-hart's convictions, arguing that empirical research studying pre-deliberation discussions in civil cases weakened the majority's rationale for finding constitutional error. See id. at 46-47, - P.3d at ----- (Bernard, J., dissenting). Next, the court of appeals held that the trial court erred by requiring Flockhart to raise and to argue his challenges for cause in the prospective jurors' presence, though it found no plain error requiring reversal. Finally, the court of appeals affirmed the trial court's order denying Flockhart's motion to disqualify the trial judge. We granted certiorari to review these rulings.1
T2 We hold that the trial court erred by giving the pre-deliberation instruction because to do so was not authorized by rule or existing law. - Cases from other jurisdictions have concluded that it is constitutional error to permit pre-deliberation discussions in criminal cases. These cases predate empirical research studying the impact of pre-delib-eration discussions in civil cases-research that weakens their rationale. Lacking pertinent empirical research conducted in erimi-nal trials and record evidence that pre-delib-eration discussions occurred here, however, we are unable to conclude that the pre-delib-eration instruction impinged upon Flock-hart's constitutional rights and instead await the results of future empirical research on this issue. Hence, we hold that that the erroneous pre-deliberation instruction in this case constituted non-constitutional trial error to be reviewed under the harmless error standard. Applying that standard, we conclude that the error was harmless. We reverse the court of appeals and remand the case to that court with directions to return it to the trial court to reinstate Flockhart's convictions.
13 Next, although the better practice is that espoused by the American Bar Association, which recommends that challenges for cause be heard outside the prospective jurors' presence, we hold that a trial court retains discretion to conduct challenges for cause in open court. Finally, we hold that the trial court did not err by denying Flock-hart's motion to disqualify the trial judge.
T4 These three issues concern discrete trial events involving unrelated pertinent facts. To analyze these issues, we provide the facts at the beginning of each analysis section.
I. Pre-deliberation Jury Instruction
1 5 We begin with the pre-deliberation jury instruction issue. We must decide whether the trial court erred by permitting pre-delib-eration discussions during Flockhart's erimi-nal trial. After recounting the facts necessary to understand this issue, we consider the propriety of permitting pre-deliberation discussions in criminal cases. Finding error, we address the appropriate standard of review and then apply that standard to this case.
A.
6 Rhoderick Flockhart was charged with distribution of marijuana and possession of eight ounces or more of marijuana, stemming *231from allegations that he had sold marijuana to a police informant and kept over nine pounds of marijuana in a building near his home.
T7 Before jury selection began, the trial court told prospective jurors that, if selected, they would be permitted to discuss the case before deliberations:
Now you will be able to discuss the case amongst yourselves, those of you that are chosen as jurors, and you may do so strictly when everyone on the jury is present. However, you are directed not to form any firm conclusions, because obviously until the case is closed, you will not have heard all the evidence.
Once selected, the trial court again instruct ed the jury that it could discuss the case as it progressed, but it cautioned the jurors to "[kJeep an open mind all the way through the trial and draw your conclusions only at the conclusion of the case." As part of its preliminary instructions, the trial court also instructed the jury on the burden of proof and the presumption of innocence. During the second day of trial, the trial court again admonished the jury to follow its previous instructions to "keep an open mind" until the conclusion of the case. Despite these instructions, there is no evidence in the record showing that any pre-deliberation discussions did in fact occur.
T8 The prosecution relied heavily on the police informant's testimony. The informant testified that he bought an ounce of marijuana, which Flockhart weighed on a kitchen scale, with catalogued bills given to him by police. The police found the kitchen scale and the catalogued bills in Flockhart's possession, and those items were admitted into evidence. Testimony at trial established that that type of kitchen seale was "commonly used for weighing drugs." The police also found nine pounds of marijuana in a building next to Flockhart's home, to which he had unfettered access from his backyard. Flock-hart's theory of defense was that the informant had set him up to curry favor with the district attorney's office, and defense counsel attempted to attack the informant's eredibility consistent with this theory. The jury apparently resolved these credibility issues in the prosecution's favor and convicted Flockhart of both charges.
19 A divided court of appeals reversed Flockhart's convictions. Relying on authority from other jurisdictions, the majority held that the pre-deliberation instruction was constitutional error subject to review under the harmless beyond a reasonable doubt standard. It then remanded the case to provide the prosecution an opportunity to make a record on that issue.
"I 10 In dissent, Judge Bernard argued that the majority's holding was based on unfounded assumptions about human behavior that have since been rebutted by empirical studies. In light of these studies, as well as jury reforms permitting pre-deliberation discussions in civil cases, Judge Bernard argued that pre-deliberation instructions are not error, let alone constitutional error.
B.
111 The People adopt Judge Bernard's arguments, contending that the court of appeals erred by relying on assumptions about juror behavior that have been rebutted by empirical research. Flockhart's argument mirrors the majority's reasoning. He contends that the pre-deliberation instruction was erroneous because it was not authorized by rule and was contrary to existing law.
112 No provision in the Colorado Rules of Criminal Procedure authorizes a trial court to issue a pre-deliberation instruction. See Crim. P. 24; People v. Preciado-Flores, 66 P.3d 155, 166 (Colo.App.2002). Colorado's pattern criminal jury instructions state that an admonition prohibiting pre-de-liberation discussion of a criminal case "should always be given" to the jury: "[Dlo not discuss the case either among yourselves or with anyone else during the course of the trial." CJI-Crim. 1:04 (1988). The pattern instructions are not law, not authoritative, and not binding on this court, but they are grounded in our longstanding practice and are regularly consulted to determine whether jury instructions are erroneous.2
*232{13 Despite this lack of authority, the People argue that we should approve pre-deliberation jury discussions given our progressive stance toward jury reform, relying on Medina v. People, 114 P.3d 845 (Colo. 2005). That case concerned whether juror questioning violates a defendant's right to a fair trial. Id. at 847. To address this issue, we analyzed the history of juror questioning, the reasons typically advanced to prohibit it, and empirical studies tending to rebut those reasons. Id. at 850-55. We held that juror questioning of witnesses does not, in and of itself and under a trial court's careful supervision, violate a defendant's right to a fair trial. Id. at 847, 857.
T 14 Our decision in Medina was based, in part, on extensive empirical research studying the effects of juror questioning, which was authorized by the Colorado Jury Reform Pilot Project and later by Crim. P.24(g).3 See id. at 847 n.1. Unlike juror questioning, pre-deliberation discussions in criminal cases have never been authorized by this court, and, hence, there is no empirical research on which we could assess the effects of pre-deliberation discussions on criminal trials. As the People point out, Colorado has embraced jury reform in other contexts, but we are wary of sanctioning pre-deliberation discussions in criminal cases without the empirical research necessary to guide our analysis. See id. at 858-55.
{15 For these reasons, we hold that the trial court erred by instructing the jury that it could discuss the case before the close of evidence.
C.
116 Finding error, we next consider the appropriate standard under which to review it. On this issue, Flockhart contends that the erroneous instruction was structural error requiring automatic reversal or, alternatively, that the error was constitutional requiring review under the harmless beyond a reasonable doubt standard. The People argue that any error was non-constitutional, and we should apply the less stringent harmless error review.
Y17 We have divided errors that occur during a criminal proceeding into two categories: structural errors and trial errors. Structural errors "affect 'the framework within which the trial proceeds." Griego v. People, 19 P.3d 1, 7 (Colo.2001) (quoting Bogdanmov v. People, 941 P.2d 247, 252-53 (Colo. 1997)). They require automatic reversal without individualized analysis of how the error impairs the reliability of the judgment of conviction. Hagos v. People, 2012 CO 63, 10, 288 P.3d 116. Trial errors are "errors in the trial process itself." Griego, 19 P.3d at 7. They do not require automatic reversal. People v. Miller, 113 P.3d 743, 749 (Colo. 2005).
[18 Thus, to identify structural error, we consider whether the error affects the "framework" of the trial rather than "the trial process itself." Griego, 19 P.3d at T7. Flockhart contends that the trial court's erroneous - pre-deliberation instruction has "framework" implications because CRE 606(b) prohibits inquiry into the jury's deliberative process, rendering the consequences of any pre-deliberation discussions "necessarily unquantifiable and indeterminate" and thus structural error. See Neder v. United States, 527 U.S. 1, 11, 119 S.Ct. 1827, 144 L.Ed.2d 85 (1999) (quoting Sullivan v. Louisiana, 508 U.S. 275, 282, 118 S.Ct. 2078, 124 L.Ed.2d 182 (1993)).
119 A trial court's erroneous pre-deliberation instruction may prejudice a defendant's constitutional right to a fair trial, but it "does not necessarily render a criminal trial fundamentally unfair." See id. at 8, 119 S.Ct. 1827. For instance, despite the instruc*233tion, pre-deliberation discussions may not take place at all, and, even if they do, such discussions may be tangential to the jurors' determination of guilt or innocence. United States v. Jadlowe, 628 F.3d 1, 20 (1st Cir. 2010). Unlike structural errors such as the complete deprivation of counsel or trial before a biased judge, the potential harm triggered by a pre-deliberation instruction does not occur unless prejudicial discussions take place. Id. Thus, a pre-deliberation instruction creates, at most, the potential for reversible error-it is not the type of instructional error that "vitiates all the jury's findings" and requires automatic reversal. See Neder, 527 U.S. at 11, 119 S.Ct. 1827 (quoting Sullvan, 508 U.S. at 281, 118 S.Ct. 2078). And, in the analogous juror misconduct context, we have reviewed instances of alleged juror misconduct for prejudice to a defendant, despite CRE 606(b). See Harper v. People, 817 P.2d 77, 86-87 (Colo.1991) (considering the jury's possible exposure to a potentially prejudicial news article); Wiser v. People, 732 P.2d 1189, 1142-43 (Colo.1987) (holding that trial courts should employ an objective test to consider whether exposure to extrancous information or influence prejudiced a defendant). Thus, we conclude that a pre-deliber-ation instruction is a trial error, not a structural error.
120 We next consider whether that trial error was constitutional. Only those errors "that specifically and directly offend a defendant's constitutional rights are 'constitutional' in nature." Wend v. People, 235 P.3d 1089, 1097 (Colo.2010) (citing Chapman v. California, 386 U.S. 18, 24-25, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967)).
1 21 The main case addressing this issue is Winebrenmer v. United States, 147 F.2d 322 (8th Cir.1945). That case involved a pre-deliberation instruction that permitted jurors to discuss the case so long as they did not form "definite fixed ideas" about the case before the close of evidence. See id. at 327. The court expressed concerns that the instruction permitted jurors to consider and to weigh evidence without having heard the trial court's instructions, that the instruction worked to shift the burden of proof to the defendant to disprove the prosecution's case, and that a juror who expresses an opinion likely will remain tethered to that opinion despite contrary evidence-such is the "pride of opinion and of consistency [that] belongs to human nature." Id. at 328 (quoting Pool v. Chicago, B. & Q.R. Co., 6 F. 844, 850 (C.C.Iowa 1881)). In light of these concerns, the court concluded that the pre-deliberation instruction violated the defendants' rights to a fair trial by an impartial jury under the Fifth and Sixth Amendments. Id. at 327, 329. Almost every court addressing this issue has echoed and adopted the Winebrenner court's rationale.4
122 Since Winebrenmer, empirical research studying pre-deliberation discussions in civil trials has weakened these concerns. The research shows that pre-deliberation discussions in civil trials encourage group decision-making, do not result in premature judgments, do not disadvantage defendants, and offer some palpable benefits-such as improving juror comprehension.5 In light of *234these studies, the Winebrenner court's reasoning, though theoretically plausible, appears questionable. See Medina, 114 P.3d at 853 (noting that a court's constitutional concerns about juror questioning were "theoretically plausible" but not verified by the scholarly work and empirical studies concerning juror questioning).
123 But these empirical studies concern the propriety of pre-deliberation discussions in civil trials, not criminal trials6 In most criminal trials, a defendant does not present a case, meaning that a jury's pre-deliberation discussions occur wholly during the prosecution's case-in-chief. In addition, most jurors in criminal cases will assume that a defendant is on trial for a reason-that some modicum of evidence exists to justify his or her presence in court. These differences, among others, may be inconsequential, but they may render pre-deliberation discussions in criminal trials constitutionally infirm. Without further research, it is impossible to know. To approve pre-deliberation discussions in criminal trials based on empirical evidence applicable only to civil trials would thus be unwise.7 Of course, "(als with any empirical analysis, these studies cannot provide a conclusive answer to the issue we confront." Id. at 854. Still, they would provide a backdrop against which we could meaningfully assess the persuasiveness of the Winebrenner rationale. Hence, we await the results of future empirical research on this issue.
124 Here, the record is silent as to whether pre-deliberation discussions occurred, let alone the substance or impact of those discussions. The empirical studies on which we would otherwise rely concern civil trials, not criminal trials, and are of little help to us here. We will not ascribe constitutional error based on "possibilities and speculation regarding factual scenarios not present in the case[ ]." See id.
[ 25 On this record, we are unable to conclude that the trial court's error specifically and directly offended Flockhart's constitutional rights. Hence, we hold that the erroneous pre-deliberation instruction in this case constituted non-constitutional trial error to be reviewed under the harmless error standard.
D.
126 Under harmless error review, reversal is required only if the error affects the substantial rights of the parties. Crim. P. 52(a); Hagos, ¶ 12. The proper inquiry to determine a harmless error question is whether the error substantially influenced the verdict or affected the fairness of the trial proceedings. Krutsinger v. People, 219 P.3d 1054, 1063 (Colo.2009).
{27 We begin with the erroneous instruction. The trial court instructed the prospective jurors that, if selected, they would be permitted to discuss the case among them*235selves before the close of evidence provided they did not come to any "firm conclusions" about the case. Once the jury was impaneled, the trial court cautioned the jurors to "[kJeep an open mind all the way through the trial and draw your conclusions only at the conclusion of the case," and it later admonished them to follow its previous instructions to "keep an open mind" throughout trial.
128 We presume that jurors follow the instructions that they receive. People v. McKeel, 246 P.3d 638, 641 (Colo.2010). Thus, in assessing the impact of the trial court's error, assuming that the jury engaged in pre-deliberation discussions in the first instance, we presume that the jurors discussed the case only when all the other jurors were present, that they kept an open mind throughout trial, and that they waited until after the close of evidence to form any "firm conclusions" about Flockhart's guilt or innocence. With this in mind, we consider whether the error substantially influenced the verdict or affected the fairness of the trial.
129 The distribution of marijuana charge was based on evidence that Flockhart sold marijuana to an informant, who had agreed to participate in Flockhart's prosecution in exchange for a plea deal in an unrelated case. At trial, the informant testified that he bought an ounce of marijuana from Flock hart with catalogued bills given to him by police. The bills were found in Flockhart's possession and served to corroborate the informant's testimony. The nine pounds of marijuana giving rise to the possession charge were found in a back room of a building next to Flockhart's home, to which Flockhart had unfettered access from his backyard. Several witnesses testified to their understanding that the back room was Flockhart's "lounge," and police conducting the search of the building testified that the building's lessee prohibited them from searching the back room because it belonged to her neighbor, Flockhart. On this record, we conclude that the erroneous pre-deliberation instruction did not substantially influence the verdict or affect the fairness of the trial proceedings.
130 Hence, we reverse the court of appeals and remand the case to that court with instructions to return it to the trial court to reinstate Flockhart's convictions. Because we have concluded that the erroneous pre-deliberation instruction was harmless, we have no need to consider the propriety of the court of appeals' remand order.
II. Challenges for Cause
{31 We next address whether the trial court erred by hearing challenges for cause in open court. After summarizing the relevant facts, we discuss the right to challenge a juror for cause. We conclude that a trial court retains discretion to hear challenges for cause in open court and then consider whether the trial court abused its discretion in this case.
A.
32 During voir dire, and in response to questions posed by the prosecution and by defense counsel, three jurors expressed doubt as to whether they could apply the presumption of innocence to Flockhart, two of those jurors expressed strong anti-marijuana sentiments, and another juror stated that her police-officer husband was "best friends" with a prosecution witness, also a police officer. When the trial court asked defense counsel whether he had challenges for cause, the following exchange took place:
THE COURT: All right, pass for cause?
[DEFENSE COUNSEL]: No. We have challenges for cause. You want to do these from the back?
THE COURT: No, on record here.
[DEFENSE COUNSEL]: In front of the jurors?
THE COURT: Yes.
133 Defense counsel then proceeded to make and to argue his challenges for cause in the presence of the prospective jurors. Defense counsel challenged the three jurors for cause because they had expressed doubt as to whether they could properly apply the presumption of innocence. The court instructed them on the presumption of innocence and asked each whether he could apply that presumption in this case. All three said *236yes, and the court denied the challenges for cause. Defense counsel then challenged the two jurors who had expressed strong anti-marijuana sentiments. The court instructed both jurors that they could not let their strong feelings override their duty as jurors to decide the case based solely on the evidence presented. Both said that they would be able to decide the case based solely on the evidence, and the court denied the challenges for cause. Last, defense counsel challenged the juror who had stated that her police-officer husband was "best friends" with a prosecution witness. The court asked the juror whether that relationship would affect her ability to weigh the evidence fairly and impartially, particularly with respect to that police officer's testimony, and the juror stated that she could serve as a fair and impartial juror to a "100 percent degree" of certainty. The court denied the challenge for cause. Defense counsel used peremptory challenges to remove the four jurors and later exhausted his allotment.
134 Recognizing a split of authority on this issue,8 the court of appeals turned to the approach espoused by the American Bar Association (ABA), which recommends that trial courts hear challenges for cause outside the presence of prospective jurors:
All challenges, whether for cause or peremptory, should be addressed to the court outside of the presence of the jury, in a manner so that the jury panel is not aware of the nature of the challenge, the party making the challenge, or the basis of the court's ruling on the challenge.
ABA Criminal Justice Trial by Jury Standards § 15-2.7(a) (3d ed. 1996). In light of this authority, the court of appeals concluded that the trial court erred but found no plain error requiring reversal.
B.
4 35 The People contend that the court of appeals erred by adopting the ABA approach and that the decision to conduct challenges for cause in open court is a matter within the trial court's discretion.
136 Due process entitles a defendant to a fair trial but not a perfect one. People v. Dunaway, 88 P.3d 619, 631 (Colo.2004). Although a defendant has a right to an unbiased jury, a defendant's rights are not necessarily violated by the possibility that a prospective juror may form a bias against him. A prospective juror with potential bias "may nonetheless sit on the jury if she agrees to set aside any preconceived notions and make a decision based on the evidence and the court's instructions." People v. Lefebre, 5 P.3d 295, 301 (Colo.2000). And, in the juror questioning context, we found no basis for the conclusion "that a juror may become antagonistic to one side should counsel object to his question." Medina, 114 P.3d at 854. These decisions reflect our reluctance to presume that every juror is irrevocably biased based on statements made during voir dire, or that every juror is inevitably biased whenever a trial court refuses to ask a juror question. In such situations, we have instead emphasized the trial court's discretionary "duty to determine the competency and ered-ibility of each juror." Id. at 856.
137 We have also recognized the trial court's discretionary authority over the conduct and the seope of the voir dire examination. We have stated that the "propriety of questions to potential jurors on voir dire is within the discretion of the trial court, and its ruling thereon will not be disturbed on appeal unless an abuse of that discretion is shown." People v. Collins, 730 P.2d 293, 300 (Colo.1986). And our Rules of Criminal Procedure recognize that trial courts possess broad discretion to place appropriate limits *237on voir dire. SeeCrim. P. 24(a)8) ("The court may limit or terminate repetitions, irrelevant, unreasonably lengthy, abusive or otherwise improper examination."); see also-Robert J. Dieter, Colorado Criminal Practice and Procedure § 17.22, at 237.
138 Consistent with this authority, we reject a per se rule compelling trial courts to hear challenges for cause outside the prospective jurors' presence, as that would require us to presume that every juror unsue-cessfully challenged for cause is inevitably biased against the party asserting the challenge. See People v. Coria, 987 P.2d 386, 391-92 (Colo.1997) ("[ClJomments [from the court] which cause disappointment, discomfort, or embarrassment to counsel in the presence of the jury, without more, rarely constitute deprivation of a fair trial."); see also Biegenwald, 524 A.2d at 137. We recognize that conducting challenges for cause in open court may result in a biased juror, and that the approach espoused by the ABA eliminates that possibility. Still, in light of our precedent, the trial court retains discretion to determine whether conducting challenges for cause in open court is appropriate depending on the unique facts and cireum-stances of each case.
139 Thus, although the better practice is that espoused by the ABA, we hold that a trial court retains discretion to conduct challenges for cause in open court. If the trial court employs this practice, it must proceed with caution. The trial court may abuse its discretion depending on the reason underlying the challenge for cause, the content of the questioning leading up to the challenge for cause, the overall tenor or contentiousness of the voir dire examination, and any other facts or cireumstances pertinent to the issue.
C.
140 Applying this standard, we consider whether the trial court abused its discretion here.
T41 Nothing in the record suggests that the challenged jurors harbored any actual bias against Flockhart as a result of the challenges for cause, and, even assuming that they did, the trial court successfully rehabilitated the challenged jurors by instructing each on his or her duty to render an impartial verdict based solely on the evidence presented at trial. The reasons underlying the challenges for cause were innocuous, common, and easily understood by the jurors, and nothing in the record suggests that the content of the questioning or the overall ten- or of the voir dire examination were such that bias would necessarily result.
T42 Hence, we conclude that the trial court did not abuse its discretion.
IIL. - Judicial Disqualification
1 43 The last issue for our review concerns whether the trial court erred by denying Flockhart's motion to disqualify the trial judge based on an appearance of partiality.
[44 Before addressing the merits of this issue, we address the People's argument that appearance-of-partiality claims are not reviewable on appeal. To support this argument, the People rely on People in the Interest of A.G., 262 P.3d 646 (Colo.2011). In that case, we held that the prejudice element of an ineffective assistance of counsel claim premised on counsel's failure to file a timely judicial disqualification motion cannot be satisfied without an allegation that the judge was actually biased, reasoning that appearance-of-partiality claims concern the reputation of the judiciary, not prejudice to the parties. Id. at 652. Contrary to the People's contention, A.G. does not stand for the broad proposition that judicial disqualification claims premised on an appearance of partiality are, as a general matter, not subject to appellate review,. A.G. merely holds that actual bias must be alleged to satisfy the prejudice prong of an ineffective assistance claim-cireumstances not present here. Thus, we may review the substance of Flock-hart's appearance-of-partiality claim, which we consider de novo. SeePeople v. Julien, 47 P.3d 1194, 1197 (Colo.2002).
1 45 After summarizing the pertinent facts, we discuss People v. Julien, in which we held that a judge's former employment with a government office, standing alone, was not an *238adequate basis for disqualification. Finding Julien instructive, we apply it to this case.
A.
T 46 Flockhart moved to disqualify the trial judge because, as a deputy district attorney seven years earlier, he had brought charges against Flockhart similar to those present here-possession of marijuana and cultivation of marijuana. In the earlier prosecution, Flockhart successfully moved to suppress evidence, after which the charges were dismissed. Having earlier served as an advocate against him, Flockhart argued that the trial judge would be unable to preside over his case in a fair and impartial manner. The trial judge remembered Flockhart's earlier prosecution but stated that he had no personal bias toward Flockhart and found no other basis that would support his disqualification.
T47 Relying on Julien, the court of appeals affirmed the trial court's order because the trial judge had no knowledge of or any connection with the investigation, preparation, or presentation of this case.
B.
1 48 Flockhart contends that Julien is distinguishable because this case involves a trial judge who, as a former prosecutor, served as an advocate against Flockhart in an unrelated criminal case. The People argue that Julien resolves this issue.
1 49 Julien involved a trial judge who had been employed by the district attorney's office while the defendant's case was pending and had taken the bench five weeks before the defendant's trial began. Id. at 1196. We held that the judge's former employment with a government office, in and of itself, was not an adequate basis for disqualification. Id. at 1198. Rather, to require disqualification on the basis of an appearance of partiality, we held that facts must exist demonstrating that the judge had personal knowledge of disputed evidentiary facts concerning the case, some supervisory role over the attorneys who were prosecuting the case, or some role in the investigation and prosecution of the case during the judge's former employment. Id.
150 As in Julien, there is no evidence in the record that the trial judge performed any role in this case, and, in contrast to Julien, the trial judge here was not employed by the district attorney's office while this case was pending-suggesting that this case presents a less troubling seenario than that in Julien.
1 51 The reality is that many judges have spent portions of their careers working for government agencies. Id. at 1199. When a former prosecutor assumes the bench as a judge, he likely will confront defendants that he has prosecuted in the past. Model Code of Judicial Conduct, Canon 3E(I) emt. at 220-22 (2004). Looking to courts in other jurisdictions that have addressed this issue, we discern "no per se rule [that] disqualifies a judge because he has prosecuted a defendant in the past." Del Vecchio v. Ill. Dep't of Corr., 31 F.3d 1363, 1375 (7th Cir.1994); see also Jenkins v. Bordenkircher, 611 F.2d 162, 167 (6th Cir.1979) (rejecting any per se rule requiring disqualification where the trial judge "has had previous contact with the defendant as a prosecutor in totally unrelated criminal charges"); Commonwealth v. Darush, 501 Pa. 15, 459 A.2d 727, 731-32 (1983) (same). Rather, the consensus appears to be that "a judge is not deemed to be disqualified from presiding over a case merely because she, or the office with which she was formerly affiliated, prosecuted a defendant on an offense that is different from the one with which defendant is presently being charged." Richard E. Flamm, Judicial Disqualification: - Recusal and Disqualification of Judges § 11.4, at 294-95 (2d ed. 2007).
$52 Like the courts in these cases, we are unwilling to adopt a per se rule requiring disqualification in every instance in which a presiding judge, as a former prosecutor, brought unrelated criminal charges against the defendant in the past. Absent facts demonstrating some material relationship between the two proceedings, or facts showing that the past prosecution is relevant to the current case, disqualification is not invariably required. See Julien, 47 P.3d at 1198; see also Leslie W. Abramson, Appearance of Impropriety: Deciding When a *239Judge's Impartiality "Might Reasonably Be Questioned", 14 Geo. J. Legal Ethics 55, 81-82 (2000).
153 Here, although both prosecutions against Flockhart involved similar marijuana charges, there was no material relationship between the two proceedings, and nothing in the record suggests that the 1999 prosecution, which occurred seven years earlier, was at all relevant to this case. Accordingly, Flockhart has not demonstrated that an appearance of partiality resulted from the trial judge's presiding over this case.
T 54 Hence, we hold that the trial court did not err by denying Flockhart's motion to disqualify the trial judge.
IV. Conclusion
4 55 For the reasons stated, we reverse the court of appeals' holding that the trial court's pre-deliberation instruction was reversible constitutional error. We remand this case to the court of appeals with directions to return it to the trial court to reinstate Flockhart's convictions
Justice EID concurs in part and concurs in the judgment in part, and Justice COATS joins in the concurrence in part and the concurrence in the judgment in part.. We granted the People's petition for certiorari and Flockhart's cross-petition for certiorari on the following issues:
1. Whether Colorado law requires challenges for cause to be raised outside the presence of the prospective jurors.
2. Whether the court of appeals erred in holding that a predeliberation instruction in a criminal case, allowing the jury to discuss the case before it is submitted to the jury, is a constitutional trial error.
3. Whether the court of appeals erred in remanding the case for a hearing where it held that a predeliberation instruction in a criminal case is constitutional trial error and it could not conclude the error was harmless beyond a reasonable doubt.
4. Whether the court of appeals erred in concluding that the trial judge did not need to recuse himself where he had, as a district attorney, recently unsuccessfully prosecuted Respondent/Cross-Petitioner on a - similar charge.
. See, eg., People v. Poe, 2012 COA 166, 110, - P.3d --, 2012 WL 4829493 (finding no error *232where jury instruction that departed from the pattern criminal instructions was nonetheless consistent with them); People v. Garcia, 2012 COA 79, 150, 296 P.3d 285 (finding no plain error in jury instruction that mirrored the pattern instruction in part because the pattern instructions "carry weight and should be considered by a trial court").
. Mary Dodge, Should Jurors Ask Questions in Criminal Cases?: - A Report Submitted to the Colorado Supreme Court's Jury System Committee (Fall 2002), http://www.courts.state.co.us./ userfiles/File/Court_Probation/Supreme_Court/ Committees/Jury_System_Standing_Committee/ dodgereport.pdf.
. The Connecticut Supreme Court held that a pre-deliberation instruction was error of constitutional dimension, reasoning that the instruction violates a defendant's constitutional right to a fair trial because it allowed jurors to discuss the evidence without the benefit of the trial court's instructions. State v. Washington, 182 Conn. 419, 438 A.2d 1144, 1147-49 (1980). Citing Winebrenner, the court also reasoned that, because "human nature" dictates that a juror who expresses an opinion is unlikely to deviate from it, the pre-deliberation instruction was constitutionally infirm. Id. at 1147-48; see also State v. McLeskey, 138 Idaho 691, 69 P.3d 111, 114, 116 (2003) (echoing the Winebrenner court's concerns before holding that an instruction allowing jurors to discuss the case before the close of evidence violated the defendant's right to a fair trial); Commonwealth v. Kerpan, 508 Pa. 418, 498 A.2d 829, 831-32 (1985) (same). In Colorado, a division of the court of appeals relied on Wirebrenner and Kerpan to reach a similar result, holding that a pre-deliberation instruction was error but not plain error requiring reversal. Preciado-Flores, 66 P.3d at 166 (summarizing the five reasons typically raised to support the argument that pre-deliberation discussions are unconstitutional).
. See Shari Seidman Diamond et al., Juror Discussions During Civil Trials: Studying an Arizona Innovation, 45 Ariz. L.Rev. 1, 31, 63-65 (2003); Valerie P. Hans et al., The Arizona Jury Reform Permitting Civil Jury Trial Discussions: The Views of Trial Participants, Judges, and Jurors, 32 U. Mich. J.L. Reform 349, 371-72 (1999); Rebec*234ca Love Kourlis & John Leopold, Colorado Jury Reform, 29 Colo. Law. 21, 22 (Feb.2000); Natasha K. Lakamp, Comment, Deliberating Juror Predeliberation Discussions: Should California Follow the Arizona Model?, 45 UCLA L.Rev. 845, 866-67, 871-73 (1998).
. See Jessica L. Bregant, Note, Let's Give Them Something to Talk About: An Empirical Evaluation of Predeliberation Discussions, 2009 U. Ill. L.Rev. 1213 (2009) (noting that no court has explicitly discussed the distinction, if any, between pre-deliberation discussions in civil trials and pre-deliberation discussions in criminal trials). One commentator has recommended permitting pre-deliberation discussions in military criminal trials because defendants in such trials do not receive the same constitutional protections as do other criminal defendants: "[Ble-cause the Sixth Amendment right to a trial by jury does not apply to the military, the prece-dential value of Winebrenner and its progeny to courts-martial practice is, arguably, nil." David A. Anderson, Let Jurors Talk: Authorizing Pre-Deliberation Discussion of the Evidence During Trial, 174 Mil. L.Rev. 92, 121 (2002).
. Only one state, Indiana, permits pre-deliberation discussions in both civil trials and criminal trials. See Ind. Jury R. 20(a)(8) (stating that the court shall instruct the jury before opening statements that the jurors, including alternates, "are permitted to discuss the evidence among themselves in the jury room during recesses from trial when all are present, as long as they reserve judgment about the outcome of the case until deliberations commence"). The rule change in Indiana was based on Arizona's rule permitting pre-deliberation discussions in civil cases and was not based on additional research of criminal trials. See Randall T. Shepard, Jury Trials Aren't What They Used to Be, 38 Ind. L.Rev. 859, 864-(2005).
. Compare Wagner v. State, 282 Ga. 149, 646 S.E.2d 676, 679 (2007) (finding no reversible error in the trial court's order denying a request "to have all challenges for cause heard outside the jurors' presence"), and State v. Hardin, 498 N.W.2d 677, 681 (Iowa 1993) (reviewing trial court's decision to conduct challenges for cause in open court for an abuse of discretion), and State v. Biegenwald, 106 N.J. 13, 524 A.2d 130, 136 (1987) (same), with Brooks v. Commonwealth, 24 Va.App. 523, 484 S.E.2d 127, 130 (1997) (concluding that counsel should be afforded the opportunity to make and to argue challenges for cause outside the prospective jurors' presence to avoid "the possibility of bias in the mind of the juror against the defendant, where bias may not have previously existed").