CHIN, J., Concurring and Dissenting.

I agree with the majority that the Court of Appeal's judgment must be reversed because the record fails to demonstrate that in resolving the administrative proceedings before the California Department of Social Services, the administrative law judge (ALJ) necessarily determined that defendant Cathy Dawn Garcia did not make misrepresentations or omissions that contributed to her receipt of overpayments. (Maj. opn., ante, at pp. 1089-1090.) However, as explained below, I disagree with the majority's decision to remand the case to the Court of Appeal for further consideration of this question. (Maj. opn., ante, at p. 1091.) Because defendant bears the burden of proving her collateral estoppel claim and the record she has provided is insufficient to meet that burden, I would hold that her collateral estoppel claim fails; she simply has not established what the majority correctly identifies as the "threshold requirements" of collateral estoppel. (Maj. opn., ante, at p. 1077.)

I also do not join the majority's discussion of whether applying collateral estoppel under the circumstances here would be consistent with public policy. The majority is reversing the Court of Appeal's judgment because of doubts that collateral estoppel's threshold requirements are met in this case. It is therefore both premature and unnecessary to decide whether, assuming the threshold requirements have been met, public policy considerations support collateral estoppel's application. Moreover, substantively, I disagree with the majority's analysis. In light of statutory developments since People v. Sims (1982) 32 Cal.3d 468 [ 186 Cal.Rptr. 77, 651 P.2d 321] (Sims), and our post- Sims decisions in Lucido v. Superior Court (1990) 51 Cal.3d 335 [ 272 Cal.Rptr. 767, 795 P.2d 1223] (Lucido), and Vandenberg v. Superior Court (1999) 21 Cal.4th 815 [ 88 Cal.Rptr.2d 366, 982 P.2d 229] (Vandenberg), were it necessary to decide the question, I would hold that public policy considerations do not support applying collateral estoppel in this case.

I. DEFENDANT HAS FAILED TO ESTABLISH THE THRESHOLD REQUIREMENTS OF COLLATERAL ESTOPPEL.

As we recently explained, collateral estoppel "applies `only if several threshold requirements are fulfilled. First, the issue sought to be precluded from relitigation must be identical to that decided in a former proceeding. Second, this issue must have been actually litigated in the former proceeding. Third, it must have been necessarily decided in the former proceeding. Fourth, the decision in the former proceeding must be final and on the merits. Finally, the party against whom preclusion is sought must be the same as, or in privity with, the party to the former proceeding. [Citations.]'" ( Pacific Lumber Co. v. State Water Resources Control Bd. (2006) 37 Cal.4th 921, 943 [ 38 Cal.Rptr.3d 220, 126 P.3d 1040].)

"The party asserting collateral estoppel bears the burden of establishing these [threshold] requirements. [Citation.]" ( Lucido, supra, 51 Cal.3d at p. 341.) Because "the law does not favor estoppels" ( People v. Frank (1865) 28 Cal. 507, 517 ( Frank)), this burden is a heavy one. As we have explained, "[c]ertainty is an essential element of every estoppel. . . ." ( Oakland v. Oakland Water Front Co. (1897) 118 Cal. 160, 221 [50 P. 277].) Thus, where a party asserts "a certain question in issue has been litigated and determined between the same parties in a previous action, it is not enough that the proposed evidence tends to show that the precise question may have been involved in such litigation." ( Emerson v. Yosemite Gold Min. etc. Co. (1906) 149 Cal. 50, 57 [ 85 P. 122].) In other words, "`[e]very estoppel must be certain to every intent, and not to be taken by argument or inference.' [Citation.] `If upon the face of a record anything is left to conjecture as to what was necessarily involved and decided, there is no estoppel in it when pleaded, and nothing conclusive in it when offered in evidence.' [Citation.]" ( Beronio v. Ventura County Lumber Co. (1900) 129 Cal. 232, 236 [ 61 P. 958].)

These rules take on special significance where the decision asserted as an estoppel may have been based on several alternative grounds. "`[I]f it appear[s] that several distinct matters may have been litigated'" in the prior action, "`upon one or more of which the judgment may have passed,'" unless the record clearly indicates "`which of them was thus litigated, and upon which the judgment was rendered,'" collateral estoppel does not apply. ( Horton v. Goodenough (1920) 184 Cal. 451, 461 [ 194 P. 34].) Thus, collateral estoppel does not apply if "it is impossible to determine from the evidence offered in support of the estoppel" which of several potentially dispositive issues the prior decision was "founded upon." ( Frank, supra, 28 Cal. at p. 516.)

Under these principles, defendant's collateral estoppel claim fails. As the majority correctly explains, the ALJ's determination here that the overpayments "were caused by administrative errors leaves open the possibility that defendant made misstatements that were a contributing cause to the overpayments." (Maj. opn., ante, at p. 1090.) Indeed, this is precisely what the trial court concluded in rejecting defendant's collateral estoppel claim. Based on the ALJ's decision, the trial court concluded that the ALJ's "causation finding" simply represents a "qualitative comparison between the things the county did and the things the defendant did" and "does not carry with it an implicit finding that there was no error in reporting." Supporting this conclusion are the ALJ's findings that "[t]he county welfare department was not fully apprised of the actual circumstances surrounding which parent had primary responsibility for the care and control of defendant's children and that "the likelihood of overpayments" would merely have been "diminished" — not eliminated — had the county welfare department properly "reviewed the [children's] living arrangement in September 1998." In light of these findings, the trial court correctly concluded the ALJ made no finding that misstatements by defendant did not contribute to the overpayments.

The majority also correctly explains that the absence in the record of the notices of action "makes it difficult to determine" precisely what issues were before and decided by the ALJ. (Maj. opn., ante, at p. 1090.) Notably, at oral argument, defendant's counsel conceded that "it's a difficult record to read" because "all" of the information and documents "weren't entered into evidence at the criminal trial." Indeed, in her brief, defendant explains that the only evidence she produced below in support of her collateral estoppel claim was a copy of the ALJ's decision. Under the governing law as set forth above, because defendant bears the burden of proving the elements of collateral estoppel, the absence from the record of the notices is her responsibility and the failure of the incomplete record to clarify the scope of the ALJ's decision requires rejection of her collateral estoppel claim. (Cf. Vella v. Hudgins (1977) 20 Cal.3d 251, 258 [ 142 Cal.Rptr. 414, 572 P.2d 28] [rejecting res judicata claim where "sparse record presented . . . fail[ed] to show either the precise nature of the factual issues litigated, or the depth of the court's inquiry"].)

I disagree with the majority's view that we should remand the case to the Court of Appeal to "consider the effect, if any, of the absences of [the notices of action] from the record. . . ." (Maj. opn., ante, at p. 1090.) As demonstrated above, by placing the burden on the party alleging collateral estoppel to prove with certainty that a particular issue was decided in the prior proceeding, and by rejecting the doctrine's application where the record fails to demonstrate which of several potentially dispositive issues the prior decision was based on, our decisions clearly specify "the effect" of the record's incompleteness (ibid.): defendant's collateral estoppel claim should be rejected. Because the majority correctly concludes that, on the record before us, the ALJ's determination leaves open the possibility that "defendant made misstatements that were a contributing cause to the overpayments" (ibid.), under the governing cases, there is nothing left for the Court of Appeal to consider. Notably, although the majority directs the Court of Appeal on remand to "consider the effect, if any, of the absences of [the notices of action] from the record" (ibid.), the majority gives the Court of Appeal no guidance as to what it is supposed to do with this information. Rather than remand for further consideration of this issue, we should simply affirm the trial court's rejection of defendant's collateral estoppel claim and its denial of her dismissal motion.

II. PUBLIC POLICY CONSIDERATIONS DO NOT SUPPORT APPLYING COLLATERAL ESTOPPEL IN THIS CASE.

However, "even where the minimal prerequisites" are established, "`"policy considerations may limit [collateral estoppel's] use where the . . . underpinnings of the doctrine are outweighed by other factors."' [Citations.]" ( Vandenberg, supra, 21 Cal.4th at p. 829.) In other words, California's collateral estoppel doctrine "has a public policy exception" ( People v. Santamaria (1994) 8 Cal.4th 903, 917, fn. 6 [ 35 Cal.Rptr.2d 624, 884 P.2d 81]) that precludes the doctrine's application, even where the threshold requirements are met, if the "policy reasons for applying collateral estoppel" are not "satisfied by the facts of [the] case." ( Sims, supra, 32 Cal.3d at p. 477.)

For several reasons, I do not join the majority's discussion of the public policy exception's application in this case. Initially, the discussion is both premature and unnecessary. As I have explained, defendant has failed to establish the threshold requirements of her collateral estoppel claim. In reversing the Court of Appeal's judgment, the majority agrees that the threshold requirements of the doctrine may not be met in this case. (Maj. opn., ante, at pp. 1089-1091.) Unless and until it is determined that defendant has established the threshold requirements, it is unnecessary to discuss whether the public policy exception to the collateral estoppel doctrine applies on the facts of this case. The majority improperly inverts the analysis, by first discussing whether the public policy exception applies, and then, as if by afterthought, discussing whether the threshold requirements have even been established. This analytical inversion cannot hide the fact that the majority's conclusion regarding the threshold question — that defendant may have failed to establish the prerequisites of her collateral estoppel claim — and its reversal of the Court of Appeal's judgment on this basis make it unnecessary to address the applicability of the public policy exception. Thus, the majority's discussion is dictum. (See People v. Pearson (1986) 42 Cal.3d 351, 358 [ 228 Cal.Rptr. 509, 721 P.2d 595] [discussion of issue that might arise on remand "was essentially dictum" where court "determined that the judgment would be reversed on other grounds"]; People v. Mendoza (2000) 23 Cal.4th 896, 914 [ 98 Cal.Rptr.2d 431, 4 P.3d 265] [where conviction was reversed because of erroneous exclusion of evidence, discussion of other issues that might arise on retrial "was not necessary to . . . case's resolution"]; Stockton Theatres, Inc. v. Palermo (1956) 47 Cal.2d 469, 474 [ 304 P.2d 7] ["discussion or determination of a point not necessary to the disposition of a question that is decisive of the appeal is generally regarded as obiter dictum"].)

More fundamentally, I disagree with the majority's analysis of the issue. At the outset, it is important to note that the majority does not make an independent analysis of whether the relevant public policy considerations support application of collateral estoppel on the facts of this case. Nor does the majority decide whether Sims was correctly decided. Instead, purporting to apply "[p]rinciples of stare decisis" (maj. opn., ante, at p. 1080), the majority simply considers whether "statutory and other changes" since Sims "warrant reconsideration of that decision. ( Id. at p. 1074.) Rather than focus almost exclusively on Sims, as does the majority, I also look to our subsequent decisions in Lucido and Vandenberg. Viewing those decisions in light of relevant statutory changes, I conclude that public policy considerations do not support application of collateral estoppel on the facts of this case.

Like the majority, I begin with Sims. There, we held that "collateral estoppel bar[red] the state from [criminally] prosecuting" a welfare recipient "for welfare fraud since she was exonerated in [an administrative fair] hearing of that charge." ( Sims, supra, 32 Cal.3d at p. 489.) In analyzing that question, we first considered whether "the technical prerequisites for applying collateral estoppel . . . were satisfied." ( Id. at p. 488.) After concluding that they were, we then considered whether "public policy considerations" supported the doctrine's application ( id. at p. 489) on "the facts of [that] case." ( Id. at p. 477.) We stated that applying the doctrine "would promote judicial economy by minimizing repetitive litigation" and would prevent "the possibility of inconsistent judgments which may undermine the integrity of the judicial system" and of "the [administrative] fair hearing process." ( Id. at p. 488.) We next explained that not applying the doctrine would work "a hardship" on a welfare recipient "who presents a successful case at the fair hearing," by requiring the recipient, "[i]n planning a budget for limited resources, . . . to take into consideration that he or she may still be required to return the benefits" that were "found" in the administrative hearing to have been "legally obtained." ( Id. at pp. 488-489.) Next, we stated that applying the doctrine "would protect [the defendant] from being harassed by repeated litigation." ( Id. at p. 489.) It "would be manifestly unfair," we declared, "[t]o subject [the defendant] to a second proceeding in which she must defend herself against the very same charges of misconduct." ( Ibid.) Finally, we stressed that "the uniqueness of the statutory scheme governing prosecutions for [welfare] fraud . . . ma[d]e application of collateral estoppel particularly appropriate in [that] case." ( Ibid.) Specifically, we cited the fact that the statutes required the state to "seek restitution by request or civil action before initiating criminal proceedings in cases involving certain categories of [welfare] fraud." ( Ibid.) By enacting this "restitution-first requirement" (maj. opn., ante, at p. 1083), we explained, "the Legislature established] a policy in favor of resolving [welfare] fraud cases outside the criminal justice system." ( Sims, supra, 32 Cal.3d at p. 489.) Only by "ignor[ing]" this legislatively established policy, we reasoned, could we hold that collateral estoppel did not apply. ( Ibid.) Based on this analysis, we concluded that collateral estoppel applied "[i]n the particular and special circumstances of [that] case." ( Ibid.)

Eight years later, in Lucido, we again considered whether "collateral estoppel principles" may be invoked "to preclude criminal trials. [Citation.]" ( Lucido, supra, 51 Cal.3d at p. 349.) There, based on a judicial finding in a probation revocation hearing that the state failed to prove an alleged criminal offense, the defendant argued that the state was collaterally estopped from criminally prosecuting him for that offense. ( Id. at pp. 340-341.) After explaining that the defendant "arguably ha[d] fulfilled the threshold requirements" of collateral estoppel ( id. at p. 341), we considered whether "the public policies underlying [the doctrine] — preservation of the integrity of the judicial system, promotion of judicial economy, and protection of litigants from harassment by vexatious litigation" — warranted its application. ( Id. at p. 343.) Regarding the first, we acknowledged that "[p]ublic confidence in the integrity of the judicial system is threatened whenever two tribunals render inconsistent verdicts. [Citation.]" ( Id. at p. 347.) However, we then explained that "[consistency . . . is not the sole measure of the integrity of judicial decisions" (ibid.), and that in this context, preserving the integrity of the judicial system requires "preserving the criminal trial process as the proper forum for determinations of criminal guilt or innocence." ( Id. at p. 350, fn. 11.) In light of this concern, we explained, we have applied collateral estoppel principles to preclude criminal trials "only when compelling public policy considerations outweighed the need for determinations of guilt and innocence to be made in the usual criminal trial setting." ( Id. at p. 349.) Regarding the second policy consideration — judicial economy — we stated that "the efficiencies of applying collateral estoppel . . . pale before the importance of preserving the criminal trial process as the exclusive forum for determining guilt or innocence as to new crimes." ( Id. at p. 351.) Regarding the third policy consideration — vexatious litigation — after noting that "[t]he essence of vexatiousness . . . is harassment through baseless or unjustified litigation," "not mere repetition," we stated: "[The defendant] does not assert that the criminal proceedings in this case are intended to harass. The public has a legitimate expectation that a person once found guilty of a crime may both be held to the terms of his probation and (if deemed appropriate by the prosecution) tried anew for any offenses alleged to have been committed during the probationary period. For this reason, it is neither vexatious nor unfair for a probationer to be subjected to both a revocation hearing and a criminal trial. The People's failure to prevail at the revocation hearing does not alone transform the otherwise permissible subsequent trial into harassment." ( Ibid.) We thus held that the finding at the defendant's probation revocation hearing did not collaterally estop the defendant's prosecution. ( Ibid.)

As the majority notes (maj. opn., ante, at p. 1084), in Lucido, we expressly explained why Sims did not require a different result. We cited Sims as a case in which "compelling public policy considerations outweighed the need for determinations of guilt and innocence to be made in the usual criminal trial setting." ( Lucido, supra, 51 Cal.3d at p. 349.) The determinative policy consideration we identified was "the `unique statutory scheme' at issue" in Sims, which expressed — through the restitution-first requirement — a "legislative determination" to "deemphasize the role of criminal trials in the overall scheme for resolution of welfare fraud cases." ( Lucido, supra, 51 Cal.3d at pp. 349-350.) We then explained: "[B]ecause [of this] legislative determination," the "concern . . . about the overall integrity of the criminal trial process as the intended forum for determinations of guilt and innocence was less at issue in Sims. . . . In the present case, by contrast, the Legislature has not indicated a preference that questions of guilt or innocence on criminal charges be litigated in revocation hearings rather than at trial. For this reason, we decline in this context to follow Sims's conclusion that preservation of the integrity of either the judicial system as a whole or the hearing process itself warrants application of collateral estoppel. Similarly, we decline to attribute as much weight in this case as we did in Sims to a need to prevent inconsistent judicial determinations." ( Ibid., fn. omitted.)

In Vandenberg, this court held that a private arbitration award, even when judicially confirmed, "may not have nonmutual collateral estoppel effect under California law unless there was an agreement to that effect in the particular case." ( Vandenberg, supra, 21 Cal.4th at p. 824, fn. omitted.) Regarding the relevant public policy considerations, the court reasoned that applying collateral estoppel in this context was not necessary either to preserve the integrity of the judicial system or to promote judicial economy. ( Id. at p. 833.) As to the former, the court explained that "because a private arbitrator's award is outside the judicial system, denying the award collateral estoppel effect has no adverse impact on judicial integrity." ( Ibid.) As to the latter, the court reasoned that "because private arbitration does not involve the use of a judge and a courtroom, later relitigation does not undermine judicial economy by requiring duplication of judicial resources to decide the same issue." ( Ibid.)

Applying Sims, Lucido, and Vandenberg in light of the current statutory and administrative scheme, I conclude that public policy considerations do not support applying collateral estoppel on the facts of this case. Under Vandenberg, "because [an ALJ's decision at a fair hearing] is outside the judicial system, denying the award collateral estoppel effect has no adverse impact on judicial integrity." ( Vandenberg, supra, 21 Cal.4th at p. 833.) Moreover, according to Lucido, the paramount policy concern in terms of the judicial system's integrity is "preserving the criminal trial process as the proper forum for determinations of criminal guilt or innocence." ( Lucido, supra, 51 Cal.3d at p. 350, fn. 11.) Lucido establishes that this concern controls unless "outweighed" by "compelling public policy considerations." ( Id. at p. 349.)

I find no such compelling policy considerations here. Concerns about judicial economy do not suffice. Under Vandenberg, "because [an administrative fair hearing] does not involve the use of a judge and a courtroom, later relitigation does not undermine judicial economy by requiring duplication of judicial resources to decide the same issue." ( Vandenberg, supra, 21 Cal.4th at p. 833.) In any event, under Lucido, any concerns about judicial economy "pale before the importance of preserving the criminal trial process as the exclusive forum for determining guilt or innocence as to new crimes." ( Lucido, supra, 51 Cal.3d at p. 351.) Nor do concerns about vexatious litigation suffice; as in Lucido, defendant here "does not assert that the criminal proceedings in this case [were] intended to harass." ( Ibid.) Moreover, to quote Lucido in this context, "[t]he public has a legitimate expectation" that a welfare recipient "may both be held to the terms" of the welfare program in an administrative fair hearing held at the recipient's request "and (if deemed appropriate by the prosecution) tried . . . for any [criminal] offenses alleged to have been committed" in connection with the program. ( Ibid.) "For this reason, it is neither vexatious nor unfair" for a welfare recipient who has requested an administrative fair hearing also "to be subjected to . . . a criminal trial." ( Ibid.) Finally, the overriding public policy consideration that, according to Lucido, allowed us to apply collateral estoppel in Sims — the Legislature's decision, expressed through the restitution-first requirement, to "deemphasize the role of criminal trials in the overall scheme for resolution of welfare fraud cases" ( Lucido, supra, 51 Cal.3d at p. 350) — no longer exists; as the parties agree and the majority "assume[s] for purposes of argument" (maj. opn., ante, at p. 1082, fn. 4), in 1984, the Legislature enacted changes to the relevant statutes that eliminated the restitution-first requirement. ( People v. Preston (1996) 43 Cal.App.4th 450, 455-461 [ 50 Cal.Rptr.2d 778].)

Indeed, the statutory and administrative scheme that now governs indicates a preference for resolution of welfare fraud cases through the criminal trial process. The Welfare and Institutions Code provides that "whenever" a person fraudulently obtains aid, that person "shall be subject to prosecution." (Welf. Inst. Code, § 11483.) The regulations of the California Department of Social Services (DSS) provide that if an investigation uncovers sufficient evidence that a welfare recipient "intentionally . . . [¶] Made a false or misleading statement, or misrepresented, concealed, or withheld facts" (DSS Manual of Policies Procedures (Feb. 8, 2005) §§ 20-300.1.11, 20-351i(1)(a) (hereafter MPP)), then a county must ask the prosecuting authority to issue a criminal complaint. ( Id., §§ 20-007.35, 20-300.21, 20-352.13.) While the prosecution is evaluating the case and "subsequent to any action taken against the accused individual by the prosecutor or court of the appropriate jurisdiction," a county may not hold an administrative hearing to determine whether the recipient engaged in such conduct. ( Id., §§ 20-300.24, 20-352.3.) Thus, whereas Sims found that application of collateral estoppel was "warranted . . . due to the unique statutory scheme which established a [legislative] preference for the noncriminal resolution of cases involving an accusation of welfare fraud" ( Sims, supra, 32 Cal.3d at p. 490), the statutory and administrative scheme now in place indicates a preference for criminal resolution of such cases. For this reason, Sims does not govern here; instead, we should follow Lucido and hold that collateral estoppel does not apply in light of the public's paramount interest in "preserving the criminal trial process as the proper forum for determinations of criminal guilt or innocence." ( Lucido, supra, 51 Cal.3d at p. 350, fn. 11.)

The majority makes several errors in concluding that there is no existing "preference for criminal resolutions" of welfare fraud cases. (Maj. opn., ante, at p. 1083.) Initially, the majority fails to consider the language of section 11483 that I have discussed above, which the People rely on in their brief. Moreover, the majority expressly declines to consider whether the "current regulatory scheme" speaks to this question, because the parties have not briefed the issue. (Maj. opn., ante, at p. 1083, fn. 5.) However, because Sims relied on the regulatory scheme in reaching its conclusion ( Sims, supra, 32 Cal.3d at pp. 480-481, 487-488), the majority, in relying exclusively on Sims, necessarily is also relying on that scheme. In my view, we should not, as the majority implicitly does, rely only on part of the regulatory scheme; we should consider all of it. If the majority believes it cannot do so on the existing record, then it should either request further briefing or decline to decide whether the public policy exception applies; it should not purport to decide that issue based on an incomplete analysis that ignores an important consideration. Indeed, in addition to its conclusion that the threshold requirements of the collateral estoppel may not be met here, the majority's view that the record is insufficient to determine the impact of the regulatory scheme further demonstrates that it is premature to decide whether the public policy exception applies.

Even were the statutes and regulations silent on the point, as explained above, Lucido is not. In this regard, although the majority purports to be concerned with "[p]rinciples of stare decisis" (maj. opn., ante, at p. 1080), it fails even to consider — much less give any weight to — the paramount public policy consideration we identified and based our decision on in Lucido: "preserving the criminal trial process as the proper forum for determinations of criminal guilt or innocence." ( Lucido, supra, 51 Cal.3d at p. 350, fn. 11.) The majority does not explain why this consideration does not apply here; instead, the majority simply ignores it. The majority also fails to consider our conclusion in Lucido that concerns about judicial economy — which Sims cited in support of its conclusion ( Sims, supra, 32 Cal.3d at p. 488) — "pale before the importance of preserving the criminal trial process as the exclusive forum for determining guilt or innocence as to new crimes." ( Lucido, supra, 51 Cal.3d at p. 351.) Nor does the majority consider Vandenberg's analysis and application of two relevant policy considerations: the judicial system's integrity and judicial economy. In purporting to follow Sims as a matter of stare decisis, the majority proceeds as if Lucido and Vandenberg, which are also entitled to respect as precedents of this court, have nothing to say about how to apply the relevant policy factors. Most notably, the majority makes no effort to resolve the tension between Lucido, which emphasized the public's interest in preserving the criminal trial process as the proper forum for determining criminal guilt or innocence, and Sims, which did not even mention this consideration.

The majority incorrectly disregards Lucido in another important respect. According to the majority, the People may avoid the collateral estoppel problem that exists under the majority's conclusion simply by pursuing a criminal case "as speedily as the County pursues the matter administratively." (Maj. opn., ante, at p. 1089.) In Lucido, we expressly rejected a similar analysis. There, after noting that "the People could avoid being collaterally estopped" by the decision in a probation revocation hearing "by prosecuting first" and "seeking revocation afterward," we stated: "As in our previous cases, . . . we refuse to mandate such a chronology. [Citation.]" ( Lucido, supra, 51 Cal.3d at p. 348, fn. 10.) Thus, in essentially mandating that a criminal prosecution be finished before completion of an administrative fair hearing, the majority again disregards Lucido.

In any event, the majority's discussion of this consideration rests on a false premise: that a county is in control of the speed with which administrative proceedings progress. It is the welfare recipient, not the county, who initiates the fair hearing process by filing a request. (§§ 10950, 10951.) Moreover, the governing regulations require that the hearing be decided within 90 days of the date the request is filed (60 days if the claim involves only food stamps), unless the recipient waives this requirement. (MPP, § 22-060.1.) Thus, the majority errs in suggesting that the speed of the administrative fair hearing process is within a county's control. The welfare recipient's control over the timing of the fair hearing process and the short regulatory time limit for deciding a fair hearing refute the majority's view that the People can easily avoid collateral estoppel's application in a criminal case simply by ensuring that prosecution is completed and a verdict rendered before an administrative fair hearing is concluded.

The majority disregards Lucido in at least one other important respect. The majority asserts that Sims merely made "observation[s]" about the restitution-first requirement that provided "nonessential support for our holding." (Maj. opn., ante, at p. 1081.) As I have explained, in Lucido, we took a decidedly different view of Sims's, discussion. ( Ante, at p. 1097.) Supporting our analysis in Lucido is our statement in Sims that refusing to apply collateral estoppel there would have required us "to ignore" the legislatively established "safeguard" established by the restitution-first requirement. ( Sims, supra, 32 Cal.3d at p. 489.) Indeed, despite its assertion, the majority elsewhere concedes that the restitution-first requirement was "one basis for our decision in Sims" and was "one factor" we considered there in applying the "multifactor analysis" that governs application of the collateral estoppel doctrine. (Maj. opn., ante, at p. 1083.) Thus, contrary to the majority's view, the issue here is not whether the restitution-first requirement was "the only ground" for our decision in Sims. (Ibid.) Rather, the issue is how to apply the governing multifactor test absent the restitution-first factor that was "one basis for our decision in Sims" (ibid.) and in light of our discussion and application of that test in Lucido and Vandenberg. Contrary to the majority's assertion, given Lucido, Vandenberg, and the relevant statutory changes, stare decisis poses no obstacle to reaching a conclusion here different from the one we reached in Sims. As the majority expressly recognizes, "`reexamination of precedent may become necessary when subsequent developments indicate an earlier decision . . . has become ripe for reconsideration. . . .'" (Maj. opn., ante, at p. 1080, quoting Moradi-Shalal v. Fireman's Fund Ins. Companies (1988) 46 Cal.3d 287, 297 [ 250 Cal.Rptr. 116, 758 P.2d 58].)

The majority's attempts to distinguish Lucido are unsuccessful. According to the majority, the facts in Lucido "did not raise the concern expressed in Sims" that allowing criminal proceedings after an ALJ's findings of no fraud and no overpayment would "leave" welfare recipients "exposed to being `harassed by repeated litigation.' [Citation.]" (Maj. opn., ante, at p. 1086.) However, although Lucido did not involve a welfare recipient, we did consider in that case the fact that applying collateral estoppel to a decision in a probation revocation hearing "would eliminate repetitive litigation" and "prevent [the defendant] from being subjected to consecutive proceedings raising the same factual allegations." ( Lucido, supra, 51 Cal.3d at p. 351.) As noted above, regarding this issue, we stated: "The essence of vexatiousness . . . is not mere repetition. Rather, it is harassment through baseless or unjustified litigation. [Citation.]" ( Ibid.) Thus, whereas Sims found that it would have been "manifestly unfair" to permit the defendant's criminal prosecution after she "successfully demonstrated her innocence" at a "fair hearing" in which the government "had an adequate opportunity . . . to prove" its case ( Sims, supra, 32 Cal.3d at p. 489), we held in Lucido that "mere repetition" is not enough to implicate the public policy concern about vexatious litigation. ( Lucido, supra, at p. 351.) Given that Sims apparently equated repetition with harassment, and did not, in considering the integrity of the judicial system and judicial economy, even mention the importance of preserving the criminal trial process as the proper forum for determining criminal guilt or innocence, the majority is simply incorrect in asserting that the "bases for our decision in Sims" other than the restitution-first requirement "remain untouched by our decision in Lucido." (Maj. opn., ante, at p. 1085.)

Also unpersuasive is the majority's view that the facts in Lucido "did not raise the concern expressed in Sims" that allowing criminal proceedings after an ALJ's findings of no fraud and no overpayment would "impose a hardship on" welfare recipients. (Maj. opn., ante, at pp. 1085-1086.) As noted above, the "hardship" we identified in Sims was that involved in requiring a welfare recipient, after establishing at a fair hearing that benefits were "legally obtained," to "plan a budget for limited resources" based on the possibility "he or she may still be required to return the benefits. . . ." ( Sims, supra, 32 Cal.3d at pp. 488-489.) In my view, forcing convicted criminals who have successfully avoided probation revocation by establishing that they did not commit a new offense to plan their lives based on the possibility they might still be convicted of and incarcerated for that offense presents an equal or greater hardship; nonetheless, that is what Lucido requires.

In any event, the facts of this case do not present the hardship at issue in Sims. Whereas the hearing officer in Sims found that the defendant had not received overpayments and was entitled to keep the money in question ( Sims, supra, 32 Cal.3d at p. 474), the ALJ here ruled that defendant did receive overpayments and had to return the extra money. Thus, in declining to apply collateral estoppel and permitting defendant's prosecution to go forward, the trial court did not threaten defendant with having to return money the ALJ ruled she could keep. As noted above, in Sims, we identified the relevant inquiry as whether "policy" considerations supported collateral estoppel's application on "the facts of [that] case" ( id. at p. 477), and we carefully limited our holding to "the particular and special circumstances of [that] case." ( Id. at p. 489.) Given these statements and the factual differences between Sims and the case now before us, the majority errs in concluding that "[p]rinciples of stare decisis" prevent us from reaching a different conclusion here. (Maj. opn., ante, at p. 1080.)

The majority also errs in asserting that this case "differs from Lucido" in that "`[probation revocation hearings and criminal trials serve different public interests'" whereas "the purposes" of administrative fair hearings and welfare fraud prosecutions "do not differ greatly." (Maj. opn., ante, at p. 1087.) The majority incorrectly indicates that the purpose of the administrative fair hearing here was "`to prove that [defendant] had fraudulently obtained welfare benefits.'" ( Ibid., italics added.) As the majority elsewhere explains, the issue at the fair hearing here was not whether defendant obtained the overpayments fraudulently, but whether they "`were the result of administrative errors or [defendant's] failure to report [her children's] absence from her home.'" ( Id. at p. 1075.) In answering this question, it was unnecessary for the ALJ also to determine whether defendant had a fraudulent intent. (See People v. Camillo (1988) 198 Cal.App.3d 981, 989, fn. 3 [ 244 Cal.Rptr. 286]; People v. Faubus (1975) 48 Cal.App.3d 1, 5-6 [ 121 Cal.Rptr. 167].) The governing statutes confirm that the DSS was not required to prove defendant acted with fraudulent intent in order to obtain reimbursement. (§ 11004.)

Indeed, under DSS regulations, the ALJ apparently lacked jurisdiction to determine whether defendant acted with fraudulent intent. The regulations provide that "intentionally" making "a false or misleading statement, or misrepresent[ing], concealing], or withh[o]ld[ing] facts" constitutes an "Intentional Program Violation" (IPV). (MPP, §§ 20-300.1.11, 20-351i(1)(a).) IPV's may be dealt with administratively only at an "administrative disqualification hearing" ( id., §§ 63-801.1.11, 63-801.231), which may be initiated only by the DSS or a county through service of notice on the welfare recipient. ( Id., §§ 22-201.412, 22-202.51, 22-315.5.) The fair hearing at issue here, which was initiated by defendant's request in response to the notices of action (maj. opn., ante, at p. 1075), was not an administrative disqualification hearing. (See MPP, §§ 22-200.2 ["Administrative disqualification hearings are distinct from" fair hearings], 22-301.2 [same].) Indeed, because a criminal action was filed in this case, an administrative disqualification hearing to determine whether defendant committed an IPV could not have been held. ( Id., §§ 20-300.24, 20-352.3.) And, because an IPV had not been established at an administrative disqualification hearing (or by a court), to the extent DSS claimed that defendant was at fault, the regulations required the claim to be "established and handled as an inadvertent household error claim." ( Id., § 63-801.23.231.) Thus, it was not within the ALJ's authority here to determine at defendant's fair hearing whether she acted with fraudulent intent.

In this regard, this case materially differs from Sims. There, the DSS alleged that the defendant had "fraudulently obtained" benefits by failing to report material information. ( Sims, supra, 32 Cal.3d at p. 473.) Under the statutory scheme that governed the defendant's fair hearing in Sims, whether the overpayments resulted from a "willful failure to report facts" or "any willfully fraudulent device," and whether the defendant had "willfully withheld information" were expressly relevant to the DSS's ability to recoup overpayments. (Former § 11004, subds. (d) (e), added by Stats. 1979, ch. 804, § 2, pp. 2767, 2768, and repealed by Stats. 1982, 1st Ex. Sess., ch. 3, § 2, p. 6890.) The defendant denied that she had failed to disclose material information and she denied that she had received benefits to which she was not entitled. Thus, the ALJ in Sims had to determine whether the defendant had acted fraudulently, and it expressly found that the DSS failed to prove she "fraudulently obtained welfare benefits." ( Sims, supra, 32 Cal.3d at p. 474.) It therefore ruled that the defendant was entitled to keep all benefits she had received and ordered the DSS to "refund any restitution payments [she] had made." ( Ibid.) Here, as already explained, the function of the fair hearing was to determine whose conduct caused the overpayments defendant received, and under the governing statute, whether defendant acted fraudulently was irrelevant to determining that question.

The purposes of administrative fair hearings and criminal prosecutions for welfare fraud appear to differ in another important respect: the nature of the causation inquiry. In the trial court, defendant's counsel asserted that the ALJ here had only "two choices" regarding causation: "either county error or client" conduct. These two choices excluded the possibility of finding that both county error and defendant's conduct were contributing factors. Counsel's argument is fully consistent with the ALJ's statement of the issue before him: "Whether the overissuance and overpayment were the result of administrative errors or [defendant's] failure to report [her children's] absence from her home." It is also consistent with the governing administrative regulations, which required the ALJ at the fair hearing to find that the overpayments were caused either by administrative error or by inadvertent household error. (MPP, §§ 63-801.211, 63-801.221.) Thus, the purpose of the fair hearing was to determine which of two mutually exclusive possible causes was the cause of the overpayments. For this reason, as the majority explains, the ALJ's decision that the overpayments were caused by administrative errors "leaves open the possibility that defendant made misstatements that were a contributing cause to the overpayments." (Maj. opn., ante, at p. 1090.) As the majority also explains, the causation inquiry in defendant's criminal prosecution for welfare fraud was different; it did not focus on determining the exclusive cause of the overpayments. In this regard, the purposes of administrative fair hearings and criminal prosecutions for welfare fraud are surely different.

In other respects, administrative fair hearings and criminal welfare fraud prosecutions "serve different public interests" in the same sense that, as we found in Lucido, "[probation revocation hearings and criminal trials serve different public interests. . . ." ( Lucido, supra, 51 Cal.3d at p. 347.) A criminal prosecution is initiated by the People to vindicate the public's "vital interest in enforcement of [its] criminal laws." ( United States v. Jorn (1971) 400 U.S. 470, 479 [ 27 L.Ed.2d 543, 91 S.Ct. 547].) That interest includes "deter[ing] the individual from committing acts that injure society" and "express[ing] society's condemnation of such acts by punishing them." (People v. Roberts (1992) 2 Cal.4th 271, 316 [ 6 Cal.Rptr.2d 276, 826 P.2d 274], italics added; see also Best v. State Bar (1962) 57 Cal.2d 633, 637 [ 21 Cal.Rptr. 589, 371 P.2d 325] ["purpose" of a "criminal proceeding" is "punishment" if "accused . . . is found guilty"].) As previously noted, an administrative fair hearing is initiated by a welfare recipient to vindicate the recipient's own private interest in challenging an "action of the county [welfare] department relating to his or her application for or receipt of public social services. . . ." (§ 10950.) It provides a "`welfare recipient with a speedy [citation] and informal [citation] means to challenge an administrative action [that] may reduce or terminate'" benefits. ( Lentz v. McMahon (1989) 49 Cal.3d 393, 402 [ 261 Cal.Rptr. 310, 777 P.2d 83].) No punishment results from an administrative fair hearing. Thus, like the probation revocation hearing at issue in Lucido, an administrative fair hearing, "despite its obvious importance to" a welfare recipient, "neither threatens" the recipient "with the stigma of a new conviction nor with punishment. . . ." ( Lucido, supra, 51 Cal.3d at p. 348.) Moreover, like "the hearing judge in a revocation proceeding," an ALJ's "fundamental role and responsibility" at an administrative fair hearing "is not to determine whether" the subject of the hearing "is guilty or innocent of a crime" (ibid.); the ALJ's responsibility is to determine whether the DSS has properly found that the welfare recipient must reimburse the agency for benefits that should not have been paid. To paraphrase Lucido, "[b]ecause the limited nature of this inquiry may not involve or invoke presentation of all evidence bearing on the underlying factual allegations, the [DSS's] failure to satisfy the lower burden of proof at the [administrative fair] hearing does not necessarily amount to an acquittal or demonstrate an inability to meet the higher criminal standard of proof. [Citation.]" ( Ibid.) For these reasons, the majority errs in asserting that "the purposes" of administrative fair hearings and welfare fraud prosecutions "do not differ greatly." (Maj. opn., ante, at p. 1087.) Accordingly, as we concluded in Lucido with respect to a probation revocation hearing, I conclude that "[preemption of trial of a [criminal] charge by [a fair hearing] decision designed to perform a wholly independent social and legal task would undermine the function of the criminal trial process as the intended forum for ultimate determinations as to guilt or innocence of newly alleged crimes." ( Lucido, supra, 51 Cal.3d at p. 349; cf. People v. Barragan (2004) 32 Cal.4th 236, 256 [ 9 Cal.Rptr.3d 76, 83 P.3d 480] [declining to apply collateral estoppel because of the public's "substantial interest in . . . `prevent[ing]' [criminal] offenders `from escaping the penalties imposed by [recidivism] statutes through technical defects in . . . proof"].)

The majority also errs in asserting that Lucido can meaningfully be distinguished by the absence in this case of "evidentiary rules [that] prohibit a welfare recipient's testimony at an administrative hearing from being introduced at a later criminal trial for welfare fraud." (Maj. opn., ante, at p. 1086.) In Lucido, we explained that such rules "guarantee the probationer the ability to present a full case at the [revocation] hearing without running the risk of prejudicing his defense at a subsequent [criminal] trial." ( Lucido, supra, 51 Cal.3d at p. 351.) In the context of administrative welfare fraud hearings, such evidentiary rules are not critical to a welfare recipient's ability fully to present his or her case. By statute, administrative fair hearings are "informal," must be conducted "in order to encourage free and open discussion by the participants," and are not constrained by the "rules of procedure or evidence applicable in judicial proceedings." (§ 10955.) DSS regulations provide that "evidence shall be admitted [at an administrative fair hearing] if it is the sort of evidence on which responsible persons are accustomed to rely in the conduct of serious affairs." (MPP, § 22-050.2.) Thus, welfare recipients can fully present their cases at administrative fair hearings through means other than their own sworn testimony. Moreover, a welfare recipient who wants to testify but who is concerned that his or her testimony may be used at a subsequent criminal trial can readily solve this potential dilemma by requesting that the fair hearing be continued until completion of the criminal prosecution. (See id., §§ 22-053 [postponements and continuances], 22-060.1 [allowing claimant to waive requirement that fair hearing be decided within 60 or 90 days of date hearing request is filed].)

In any event, the majority's focus on this issue is a red herring because the dilemma exists even under the majority's view that collateral estoppel may, as a matter of public policy, apply in this context. Welfare recipients who lose at administrative fair hearings cannot assert collateral estoppel and thus may have their hearing testimony used at any subsequent criminal trial. Of course, at the time they must decide whether to testify at the administrative hearing, welfare recipients cannot know whether they will win or lose. Thus, at the time the decision whether to testify must be made, even under the majority's view that collateral estoppel applies if they win, welfare recipients face the risk they will lose and that their testimony will therefore be admissible against them at a subsequent criminal trial. Therefore, in this context, even under the majority's view, welfare recipients trying to decide whether to testify at administrative hearings must still consider the possibility that their testimony may later be used against them in criminal proceedings. Because the concern the majority discusses exists whether or not collateral estoppel potentially applies, the majority errs in relying on this consideration as a basis for reaching a policy conclusion here different from the one we reached in Lucido.

Indeed, contrary to the majority's analysis, under Vandenberg, the most meaningful distinction between this case and Lucido establishes that it would be even more appropriate — not less — to apply collateral estoppel's public policy exception in this case than it was in Lucido. The probation revocation hearing in Lucido was a judicial proceeding decided by a judge of a "justice court." ( Lucido, supra, 51 Cal.3d at p. 339.) The fair hearing in this case was an administrative proceeding decided by a nonjudicial officer. As previously explained, under Vandenberg, that administrative fair hearings are "outside the judicial system" and do "not involve the use of a judge and a courtroom" supports the conclusion that, as a matter of public policy, we should not apply collateral estoppel in this case. ( Vandenberg, supra, 21 Cal.4th at p. 833.)

The majority's stated reason for disregarding Vandenberg is inadequate. According to the majority, Vandenberg involved "a private arbitration" that "effectively `bypass[ed] the judicial system,'" whereas this case involves "a matter of public concern" dealt with in "`a judicial-like adversary proceeding' [citation] that [was] conducted by a state entity, the DSS, whose administrative law judges must follow California law and render decisions subject to judicial review. [Citation.]" (Maj. opn., ante, at p. 1086, fn. 6.) In making this assertion, the majority once again ignores prior precedent of this court, which characterizes "private arbitration proceedings" as "`quasi-judicial' proceedings . . . that are functionally equivalent to court proceedings. [Citation.]" ( Moore v. Conliffe (1994) 7 Cal.4th 634, 645 [ 29 Cal.Rptr.2d 152, 871 P.2d 204].) The majority also ignores the fact that in Vandenberg, the arbitration "took place before a retired federal judge," the parties conducted "[f]ormal discovery," "the transcribed proceedings included representation by counsel, and extensive evidence, briefing, and argument," the arbitrator issued "a lengthy and detailed decision," and the arbitration award was judicially "confirmed by a superior court judgment" ( Vandenberg, supra, 21 Cal.4th at p. 826), which gave it "the same force and effect as . . . a judgment in a civil action of the same jurisdictional classification" and made it enforceable "like any other judgment of the court in which it [was] entered. . . ." (Code Civ. Proc., § 1287.4.) By contrast, in this case, the administrative proceeding never involved a court in any capacity. Because the ALJ's decision here was entirely "outside the judicial system" and the fair hearing did "not involve the use of a judge and a courtroom" ( Vandenberg, supra, 21 Cal.4th at p. 833), Vandenberg's discussion of these considerations in connection with arbitration fully applies, whether or not, as the majority asserts, the fair hearing was "`judicial-like'" and involved "a matter of public concern." (Maj. opn., ante, at p. 1086, fn. 6.)

Finally, for several reasons, the majority's assertion that the Legislature, in amending the welfare fraud statutes in 1984, "did not contemplate abrogating Sims" (maj. opn., ante, at p. 1087) is unpersuasive. First, it is largely beside the point. As we have explained, the result in Sims was a judicial policy decision that was "informed by the `unique statutory scheme' at issue there. [Citations.]" ( Gikas v. Zolin (1993) 6 Cal.4th 841, 851 [ 25 Cal.Rptr.2d 500, 863 P.2d 745].) Thus, the primary question here is not whether the Legislature intended to abrogate Sims, but whether we should make a different policy decision in this case because, among other things, the "`unique statutory scheme'" that "informed" our policy decision in Sims (ibid.) has been materially changed.

Second, the majority's analysis is inconsistent with our prior decisions. The majority asserts that we must presume the Legislature was aware of Sims and would have specified that administrative decisions do not preclude criminal prosecutions for welfare fraud had it wanted to invalidate Sims. (Maj. opn., ante, at p. 1088.) We rejected a very similar argument in Estate of Kachigian (1942) 20 Cal.2d 787 [ 128 P.2d 865]. At issue there was our construction of the probate homestead statutes, which we had based on the homestead statutes that apply before death. ( Id. at pp. 788-790.) The question we faced was whether a legislative amendment to the latter affected the former. ( Ibid.) Offering an analysis very similar to the majority's, the respondent in Kachigian, who contended that the amendment did not effect a change, argued: "[I]t must be presumed that the Legislature was familiar with the former decisions and that in failing to change the probate statutes it must have intended to leave the probate law unchanged." ( Id. at p. 790.) We rejected the argument, stating: "[I]f we presume that the Legislature was familiar with the former decisions, it would seem improper for us to presume that it was unaware of their basic reasoning and hold that it did not realize the probate rule must change if the basis therefor were changed." ( Ibid.) Similarly, here, "if we presume that the Legislature was familiar with [Sims], it would seem improper for us to presume that [the Legislature] was unaware of [Sims's] basic reasoning and hold that [the Legislature] did not realize the [Sims] rule must change if [a] basis therefor were changed." ( Ibid.) Thus, I disagree with the majority's view that the Legislature, in repealing the statutory requirement that informed our decision in Sims, "did not contemplate abrogating Sims." (Maj. opn., ante, at p. 1087.)

As the preceding analysis demonstrates, ultimately, in concluding that Sims does not "warrant reconsideration" (maj. opn., ante, at p. 1074), the majority is focusing on the wrong question. As noted above, in Sims, we considered whether "policy" considerations supported collateral estoppel's application on "the facts of [that] case" ( Sims, supra, 32 Cal.3d at p. 477), and we expressly limited our holding to "the particular and special circumstances of [that] case." ( Id. at p. 489.) As also explained above, the facts and circumstances of the case now before us are different: the restitution-first requirement that was a "basis for our decision in Sims" (maj. opn., ante, at p. 1083) no longer exists, the ALJ ruled that defendant must repay the overpayments she received, and the question of whether defendant intentionally made misrepresentations or omissions was not at issue in the administrative proceeding. Thus, the question we must answer here is not, as the majority posits, whether to reconsider Sims, but is whether to extend that decision to the different facts and circumstances of this case. More precisely, under Lucido, a precedent of this court that the majority simply ignores in relevant part, the question we must answer is whether "compelling public policy considerations outweigh the need for determinations of guilt and innocence to be made in the usual criminal trial setting." ( Lucido, supra, 51 Cal.3d at p. 349.) Because I find no such consideration and the majority fails to identify one, and because the majority offers no meaningful basis for distinguishing Lucido and concluding that although collateral estoppel does not apply after probation revocation hearings, it may apply after administrative fair hearings, were it necessary to decide the question here, I would hold that public policy considerations do not support applying collateral estoppel in "the particular and special circumstances of this case." ( Sims, supra, 32 Cal.3d at p. 489.)

Corrigan, J., concurred.