OPINION

WERDEGAR, J.

Prosecutors are public fiduciaries. They are servants of the People, obliged to pursue impartially in each case the interests of justice and of the community as a whole. When conflicts arise that compromise their ability to do so, they can and should be recused. But defendants bear the burden of demonstrating a genuine conflict; in the absence of any such conflict, a trial court should not interfere with the People's prerogative to select who is to represent them.

In this and a companion case, Hollywood v. Superior Court (2008) 43 Cal.4th 721, we consider the extent to which involvement in literary and cinematic endeavors may give rise to conflicts requiring recusal. Here, the lead prosecutor moonlighted as a novelist, writing a fictional account of a heroine prosecutor's decision whether to try a rape case involving an intoxicated victim. The novel was published shortly before her scheduled prosecution of petitioner Massey Harushi Haraguchi for the rape of an intoxicated victim. In this case, as in Hollywood, the trial court found no conflict. In opinions issued the same day, the same Court of Appeal — relying in part on the unusual and distinctive facts of these cases — exercised its independent judgment and in both cases reversed and ordered recusal. Thus, we must also consider the standard for reviewing a trial court's decision finding, or rejecting, the existence of a disqualifying conflict.

We reaffirm our long-standing rule that recusal motions are reviewed under a deferential abuse of discretion standard. Furthermore, we reverse the Court of Appeal here based on its failure to grant appropriate deference and based on the presence in the record of evidence sufficient to support the trial court's conclusion that no disqualifying conflict existed and no unlikelihood of a fair trial had been proven.

FACTUAL AND PROCEDURAL BACKGROUND

On September 14, 2005, an information was filed charging Haraguchi, inter alia, with rape of an intoxicated person. (Pen. Code, § 261, subd. (a)(3).)

On April 25, 2006, Haraguchi moved for recusal of Deputy District Attorney Joyce Dudley and the Santa Barbara County District Attorney's Office based on Dudley's publication of a novel, Intoxicating Agent, in January 2006, as well as her subsequent promotion of the book. Haraguchi alleged, inter alia, that (1) the book contained a lengthy fictional account of the rape of an intoxicated person; (2) the fictional trial, like Haraguchi's, was to begin around April 2006; (3) a character in Dudley's novel bore a resemblance to Haraguchi; and (4) the facts of the fictional rape in other respects mirrored another unrelated case in which Dudley had obtained only a hung jury. Haraguchi further contended that Dudley was marketing the book locally, selling it in at least two local bookstores, conducting a book signing at one of the bookstores and at the Women's Center of the University of California at Santa Barbara, and doing an April 4, 2006, interview with a Santa Barbara television station about the book. In addition, the Santa Barbara Independent published a favorable review of the novel.

Haraguchi contended Dudley's publication and promotion of her book was influencing how she handled his case. According to Haraguchi's counsel, "[w]hen I took over the Haraguchi case from prior counsel, Ms. Dudley's first unsolicited remarks to me were . . . that no other prosecutor at the DA's office would take a case like Haraguchi, but that she could win it." Counsel argued that Dudley's remarks were similar to views in the novel expressed by the protagonist, Santa Barbara County District Attorney Jordon Danner: "Jordon was positive another D.A. wouldn't want this case; and, she knew the victim would feel betrayed if she handed it off." (Dudley, Intoxicating Agent (2006) p. 57 (Intoxicating Agent).) Counsel further declared: "From the first time the undersigned talked to Ms. Dudley about this case, she said it would not settle. She has refused to enter into any settlement negotiations and has summarily informed the court that the case will go to trial." Based on this, Haraguchi contended Dudley's interest in promoting her book was compromising her willingness to seek justice impartially.

In opposition, Dudley disputed these allegations. She submitted a declaration denying that Intoxicating Agent was based in any way on the Haraguchi case, that its publication or publicity were in any way coordinated with the Haraguchi case, that Haraguchi's counsel had accurately represented her remarks, or that her decisions had in any way been shaped by the book's publication.

The trial court denied Haraguchi's motion, finding there was no conflict warranting recusal. It concluded publication of the book around the same time as Haraguchi's trial was coincidental; the rape case described in the book was unrelated to Haraguchi's case; the alleged physical resemblance of a character (not the fictional rapist) to Haraguchi was not prejudicial; and, to the extent there might be any conflict, it was not so grave as to render it unlikely Haraguchi would receive a fair trial.

The Court of Appeal granted Haraguchi's petition for a writ of mandate. Declining to defer to the trial court because the trial court "had no precedent to guide it," the Court of Appeal held recusal was required as a matter of law. It concluded Dudley's views of the justice system, as reflected in the novel, were so one-sided as to raise a reasonable possibility she would not exercise her discretion evenhandedly, and her interest in promoting her book presented a conflict so great it was unlikely Haraguchi could receive a fair trial. However, the Court of Appeal declined Haraguchi's further request to recuse the entire office.

We granted review to consider both the standard of review and its application to a prosecutorial recusal motion in these circumstances.

DISCUSSION

I. Standards for a Motion to Recuse

Section 1424 sets out the standard governing motions to recuse a prosecutor: such a motion "may not be granted unless the evidence shows that a conflict of interest exists that would render it unlikely that the defendant would receive a fair trial." ( Id., subd. (a)(1).) The statute "articulates a two-part test: `(i) is there a conflict of interest?; and (ii) is the conflict so severe as to disqualify the district attorney from acting?'" ( Hambarian v. Superior Court (2002) 27 Cal.4th 826, 833 [ 118 Cal.Rptr.2d 725, 44 P.3d 102], quoting People v. Eubanks, supra, 14 Cal.4th at p. 594.)

We uniformly have held that a motion to recuse is directed to the sound discretion of the trial court, and its decision to grant or deny the motion is reviewed only for an abuse of discretion. ( People v. Vasquez (2006) 39 Cal.4th 47, 56 [ 45 Cal.Rptr.3d 372, 137 P.3d 199]; Hambarian v. Superior Court, supra, 27 Cal.4th at p. 834; People v. Eubanks, supra, 14 Cal.4th at p. 594; People v. Superior Court (Greer), supra, 19 Cal.3d at p. 269.) The abuse of discretion standard is not a unified standard; the deference it calls for varies according to the aspect of a trial court's ruling under review. The trial court's findings of fact are reviewed for substantial evidence, its conclusions of law are reviewed de novo, and its application of the law to the facts is reversible only if arbitrary and capricious.

The Court of Appeal in this case gave no deference to the trial court's findings of fact or its application of the law to those facts. It offered two justifications for this departure: (1) this case involved novel circumstances and was thus a matter of first impression; and (2) the Court of Appeal had an independent interest in eliminating errors before trial. Neither persuades us to create an exception to the choice of the abuse of discretion standard.

With respect to this case being a matter of first impression, we note that virtually every case is, to a greater or lesser degree, a matter of first impression. The difference between each new set of facts and those that previously have been ruled upon may be small and immaterial or large and momentous. Where on the continuum a new set of facts lies is to some extent in the eye of the beholder; a court of a mind to reverse may always point to those elements of a case that it views as distinguishing and on that basis assert the issue is a matter of first impression. The Court of Appeal's proposal that cases of "first impression" should receive independent review thus offers no meaningful boundaries and wholly undermines the trial court's role in evaluating recusal motions in the first instance.

As an alternative ground for exercising independent review, the Court of Appeal pointed to its interest in policing conflicts of interest and addressing potential errors at the earliest opportunity, rather than on direct review or in habeas corpus proceedings. Its concerns are valid; we have acknowledged that pretrial recusals serve the important function of "avoid[ing] conflicts that might lead ultimately to due process violations and hence to reversals or mistrials." ( People v. Vasquez, supra, 39 Cal.4th at p. 59.) These concerns justify vesting trial courts with broad discretion to protect against procedural unfairness by ordering pretrial recusals, as we emphasized in Vasquez and in People v. Superior Court (Greer), supra, 19 Cal.3d at pages 264-265. They may also justify allowing the decision to grant or deny recusal to be reviewed pretrial through mandamus, as it was here.

The Court of Appeal's concerns do not, however, support a change in the standard of review. The assertion that pretrial review should be de novo rests on the unspoken assumption that independent review will reduce the rate of error — that appellate courts given a free hand to weigh the evidence and disregard trial court findings will reverse erroneous rulings and eliminate error more often than they reverse correct rulings and thereby introduce error. That assumption is unfounded. We review rulings on motions to recuse only for abuse of discretion precisely because trial courts are in a better position than appellate courts to assess witness credibility, make findings of fact, and evaluate the consequences of a potential conflict in light of the entirety of a case, a case they inevitably will be more familiar with than the appellate courts that may subsequently encounter the case in the context of a few briefs, a few minutes of oral argument, and a cold and often limited record. (See People v. Roldan, supra, 35 Cal.4th at p. 688 [abuse of discretion standard "reflects the trial court's superior ability to consider and weigh the myriad factors that are relevant to the decision at hand"]; People v. Conner, supra, 34 Cal.3d at p. 149 [trial court is in the best position to assess the impact of events giving rise to possible prosecutorial conflicts]; People v. Jenan (2006) 140 Cal.App.4th 782, 793 [ 44 Cal.Rptr.3d 771] [abuse of discretion standard applies because the trial court is in the best position to make relevant conflict determinations].) Nothing suggests to us that de novo review of pretrial recusal motions would increase the accuracy of such determinations. As the trial court has the superior vantage point, the abuse of discretion standard applies.

II. Application to Authors/Attorneys

We turn to the application of the recusal statute to the lead prosecutor in Haraguchi's case. As noted, section 1424 sets out a two-part test for determining whether recusal is appropriate. Under the first part, a court must determine whether a conflict exists, that is, whether "the circumstances of a case evidence a reasonable possibility that the DA's office may not exercise its discretionary function in an evenhanded manner." ( People v. Conner, supra, 34 Cal.3d at p. 148; accord, People v. Eubanks, supra, 14 Cal.4th at p. 594.) If such a conflict exists, the court must further determine whether the conflict is "`"so grave as to render it unlikely that defendant will receive fair treatment during all portions of the criminal proceedings."'" ( People v. Vasquez, supra, 39 Cal.4th at p. 56, italics omitted.) Thus, the first half of the inquiry asks only whether a "reasonable possibility" of less than impartial treatment exists, while the second half of the inquiry asks whether any such possibility is so great that it is more likely than not the defendant will be treated unfairly during some portion of the criminal proceedings.

A. Existence of a Conflict

The trial court concluded there was no "conflict that would justify recusal of Ms. Dudley." It based that conclusion on a series of factual findings: "The publication of her book appears to be coincidental to Mr. Haraguchi's circumstances. The circumstances related in her book factually don't appear to relate to Mr. Haraguchi's circumstances, and the fact that there may be similarities to some other case that Ms. Dudley tried in the past doesn't establish any conflict as far as Mr. Haraguchi's case is concerned. Whatever Ms. Dudley's feelings about the results of that prior prosecution are, those feelings would be present with or without the existence of the book and don't create any conflict of interest."

In reversing the trial court and concluding that Dudley had to be recused as a matter of law, the Court of Appeal found two alleged conflicts. First, it reasoned that publishing a book linked to the Haraguchi case gave Dudley a financial incentive to prosecute the case, as the publicity might increase her book sales. Second, it concluded the fictional prosecutor Jordon Danner's views could be imputed to Dudley (and, more generally, that the views about the criminal justice system as a whole reflected in the book's interior world could be imputed to Dudley), and that those views rendered it unlikely Haraguchi could receive a fair trial.

In reasoning that Dudley had a disqualifying financial conflict of interest because of links between Intoxicating Agent and the Haraguchi case, the Court of Appeal disregarded the trial court's findings that Intoxicating Agent was not factually based on Haraguchi, its publication was not intentionally timed to coincide with Haraguchi, and its publicity was not linked to Haraguchi. Each finding was supported by substantial evidence. In a declaration, Dudley asserted under oath that Intoxicating Agent's publication was not timed to coincide with Haraguchi's trial. Her declaration likewise asserted that she did not base the plot on the Haraguchi case, and a comparison of the book with the Haraguchi record fully supports that assertion.

The trial court was entitled to credit Dudley's declaration. Likewise, review of the evidence in the record confirms the trial court's conclusion that Haraguchi failed to establish any interconnection between his case and the promotion of Intoxicating Agent; none of the promotional materials submitted by Haraguchi in support of his recusal motion allude in any way to his case. We thus are required to accept, as was the Court of Appeal, the findings, supported by substantial evidence, that "[t]he publication of her book appears to be coincidental to Mr. Haraguchi's circumstances" and that "[t]he circumstances related in her book factually don't appear to relate to Mr. Haraguchi's circumstances."

The trial court's findings that the rape case in Intoxicating Agent was not based on and was factually dissimilar to the Haraguchi prosecution are significant. Because there was no meaningful factual connection between the two, publication of the book created little incentive for Dudley to handle the Haraguchi prosecution any differently than she otherwise would have. That is, while a prosecutor's literary career might benefit generally from publicity attendant on successful prosecutions or plea bargains, there is little reason to conclude such a second career would have distortive effects on the incentives to try, settle, or dismiss any particular unrelated case. Dudley would not gain from insisting on pressing forward with a case whose merits might otherwise suggest a different disposition, rather than settling or dismissing it and devoting prosecutorial resources to another case with a better chance for a favorable outcome. Thus, whatever financial incentives her novel might have created for Dudley, those incentives were not likely to alter how she handled the Haraguchi case. (Cf. People v. Eubanks, supra, 14 Cal.4th at p. 598 [holding that financial incentives tied to a particular case that might skew the charging or plea bargaining decisions in that case create a cognizable conflict].)

As a further basis for recusal, the Court of Appeal concluded Dudley's personal views about the criminal justice system, as reflected by Intoxicating Agent, would preclude Haraguchi from receiving a fair trial. In the world of Intoxicating Agent, as the Court of Appeal interpreted it, defendants are villainous, defense attorneys are manipulative schemers, and prosecutors are heroes. Moreover, the Court of Appeal directly attributed the views of the fictional prosecutor, Jordon Danner, to Dudley.

However, as the trial court correctly recognized, Ms. Dudley's views about her role as a prosecutor, and her ability to fairly and impartially represent the interests of the People, are unaffected by and largely independent of whether a fictional character in a novel she wrote might not embody the prosecutorial ideal in the eyes of a reviewing court. Dudley is not Danner; Danner is not Dudley. Whatever imbalanced views Danner may hold, Dudley may not automatically be charged with them. The same is true of the world view presented in Intoxicating Agent, which is, after all, a work of fiction. That world view at best offers only a distorted lens through which to perceive what Dudley's views might be. Protagonists in a novel need not be paragons of impartiality to permit their creators to fairly prosecute criminals; likewise, the landscape of a novel need not precisely mirror the balanced and fair justice system our society aspires to in order to allow its author to participate in that system.

The Court of Appeal deconstructed the author's acknowledgments in order to attribute the book's entire point of view to her. The acknowledgments, which indicate Dudley tried to prepare a "trustworthy" novel, do not establish that every opinion in the book mirrors Dudley's. Even if they did, those opinions would demonstrate a conflict only if they were so extreme as to somehow establish a reasonable possibility Dudley would be unable to exercise her discretion fairly. As the trial court could reasonably conclude, they are not.

Moreover, the Court of Appeal misinterpreted the views of Dudley's fictional alter ego. The court highlighted the following statement by Danner: "[I]t was well known throughout the legal community that a negative outcome on these kinds of [rape] cases had a trickle-down effect, resulting in other potential assailants believing if they preyed upon an intoxicated victim, they could get away with rape." (Intoxicating Agent, supra, at p. 68.) From this, the Court of Appeal concluded Dudley was a conflicted prosecutor who somehow might be less willing to accept a plea bargain to a lesser charge — a negative outcome, in the Court of Appeal's eyes — than some other prosecutor. Putting aside for the moment the larger point, that the fictional Danner's views have limited relevance to a decision whether to recuse Dudley, it is apparent from the novel's preceding sentence that the "negative outcome[s]" referred to were "not guilty verdicts, or even hung juries." A prosecutor interested in protecting the interests of the community as a whole might well factor in, in deciding whether to try a particularly difficult case, what impact a verdict short of guilty might have on the safety and well-being of that community. Nothing in the fictional character's views suggests an unwillingness to offer or accept a plea bargain that might otherwise be warranted. To the contrary, the fictional Danner recognized the need to "go through all the evidence, critically, for the [victim's] sake, as well as for the rest of the community, and if [she] didn't feel she could prove this defendant's guilt to the standard of `beyond a reasonable doubt,' she'd have a duty to accept [defense counsel's] offer" to plead guilty only to sexual battery. (Intoxicating Agent, supra, at p. 68.)

The trial court's role, and the Court of Appeal's and ours, is to examine the record for evidence of a disqualifying conflict, not to act as literary critic. That a prosecutor may pursue an independent writing career does not alone create a conflict with the public interest and disqualify her from future prosecutions, absent proof her writings create a material conflict in a particular case. Here, on the record before it, the trial court permissibly could conclude there was none.

In this court, Haraguchi relies on these same grounds in asserting a conflict: that Dudley wrote a book (conceded) related to his own case (contradicted by the trial court's findings); that the Santa Barbara County District Attorney's Office was portrayed and, as suggested by the acknowledgements, involved in the book's preparation (not alone a basis for a conflict); that Intoxicating Agent gave a one-sided view of the criminal justice system (likewise not alone a basis for a conflict); and that the release of the book to coincide with Haraguchi's trial gave Dudley an incentive to handle that trial differently (contradicted by the trial court's findings). While it certainly would not have been an abuse of discretion for the trial court to conclude that Dudley's authorship and promotion of Intoxicating Agent created a reasonable possibility she might handle the discretionary duties of her office differently in Haraguchi's case, neither was it an abuse of discretion to conclude, as the trial court did here, that there was no conflict. Explicitly part of the trial court's ruling was an assessment that this book was sufficiently low profile, and Haraguchi's case sufficiently low profile, that the book's publication would not affect Dudley's judgment and exercise of discretion in prosecuting Haraguchi.

B. Gravity of Any Conflict

Out of an appropriate abundance of caution, the trial court also considered the second half of the recusal test — whether any conflict was so grave it would render a fair trial unlikely (§ 1424, subd. (a)(1)) — and concluded Haraguchi had failed to establish this element as well: "I don't think there's any evidence of that. It has not been demonstrated or established that any publicity related to Ms. Dudley's book has been so extensive or interlinked with Mr. Haraguchi's case that he would be unlikely to receive a fair trial."

Substantial evidence supports the trial court's conclusion that the publicity attending Intoxicating Agent was not so great as to render a fair trial unlikely. Haraguchi submitted evidence of only a smattering of local public appearances by Dudley in support of her novel. Essentially self-published, the book has not been widely purchased. Even if one were to reach a different conclusion on the question of conflict, and hold as the Court of Appeal did that Intoxicating Agent's publication created at least a reasonable possibility Dudley's discretionary decisions might be influenced, the minimal publicity and sales indicate any financial incentives she might have to act differently are de minimis and do not establish the likelihood of unfair treatment the statute requires.

In addition, the trial court found any potential taint to the jury pool from the promulgation of the views in Intoxicating Agent could be handled through a sequestered voir dire. It proposed that in voir dire, attorneys could inquire whether prospective jurors were familiar with any of the writings of either the prosecutor or defense attorney (who it appears had various writings of his own); if any answered yes, further questioning to deal with any possible bias could be conducted out of the presence of the rest of the jury pool. This is precisely the sort of discretionary trial-management decision we vest trial courts with, and the trial court did not err in concluding this sort of potential prejudice could be dealt with in that fashion.

In defending its decision to require recusal, the Court of Appeal offered its view that allowing Dudley to prosecute Haraguchi would be "unseemly." It acknowledged that unseemliness alone is not a basis for recusal. Section 1424 "does not allow disqualification merely because the district attorney's further participation in the prosecution would be unseemly, would appear improper, or would tend to reduce public confidence in the impartiality and integrity of the criminal justice system." ( People v. Eubanks, supra, 14 Cal.4th at p. 592; see also People v. Neely, supra, 70 Cal.App.4th at pp. 777-779; People v. McPartland (1988) 198 Cal.App.3d 569, 574 [ 243 Cal.Rptr. 752].) Only an actual likelihood of unfair treatment, not a subjective perception of impropriety, can warrant a court taking the significant step of recusing an individual prosecutor or prosecutor's office. Nevertheless, the Court of Appeal deemed it unseemly that Intoxicating Agent presents a one-sided view of the criminal justice system and that Dudley would take advantage of her office for financial gain by using the Santa Barbara County District Attorney's Office as the backdrop for her novel. But the first of these does not, as we have discussed, create any semblance of a conflict, and whatever one might think of the propriety of the second of these, it likewise does not lead to a "reasonable possibility" that Dudley would fail to exercise the discretionary duties of her office in a fair and evenhanded manner. Consequently, whether or not the Court of Appeal was justified in disapproving Dudley's perceived misuse of her office, recusal is not the remedy.

In upholding the trial court's exercise of discretion, we do not condone actions that place a prosecutor's literary career ahead of, or at odds with, her fealty to the fair and evenhanded pursuit of justice and the community interest. Writers are often encouraged to "write what they know," but the prosecutor who follows that advice in ways that touch on pending matters may compromise her ability to carry out her duties to represent the People and to seek justice impartially. In this case, the trial court concluded based on substantial evidence that no such disabling conflict had arisen. Giving the trial court the deference due its judgment, we conclude that determination should not have been disturbed.

DISPOSITION

For the foregoing reasons, we reverse the Court of Appeal's judgment and remand this case for further proceedings consistent with this opinion.

George, C. J., Kennard, J., Baxter, J., Chin, J., Moreno, J., and Corrigan, J., concurred.