Opinion
BROWN, J.The Penal Code provides that inmates in county jails and state prisons may have their sentences reduced as a reward for their conduct, including work and good behavior. The rate at which inmates accrue credit depends on numerous factors, including whether the confinement is presentence or postsentence. A jury convicted petitioner, her conviction was reversed, and she then pleaded guilty. She now contends her state prison confinement prior to the reversal must be characterized as presentence for credit accrual purposes.
In People v. Buckhalter (2001) 26 Cal.4th 20 [108 Cal.Rptr.2d 625, 25 P.3d 1103] (Buckhalter), we held that a petitioner who is serving a state prison sentence and is remanded for resentencing retains postsentence status for credit accrual purposes. (Id. at pp. 40-41.) We expressly declined to “consider the proper credit treatment of one who spends time in custody after his convictions have been reversed on appeal, thus setting the entire matter at large.” (Id. at p. 40, fn. 10.) We now face the question left unanswered in Buckhalter. We conclude petitioner’s prereversal prison time ought not be viewed as presentence custody, and her credit accrual should be calculated in accordance with her ultimate postsentence status.
I. Proceedings Below
Petitioner was taken into custody on February 11, 1995. A jury convicted her of petty theft with priors (Pen. Code, § 666),1 and determined she had two prior strikes within the meaning of sections 667 and 1170.12. On July 9, 1996, the trial court sentenced petitioner to a state prison sentence for 25 years to life pursuant to the three strikes law. On July 9, 1999, petitioner obtained a reversal of her conviction on habeas corpus review due to her receiving ineffective assistance from trial counsel. On August 19, 1999, *32petitioner pleaded guilty to the underlying charge, and the trial court dismissed one of her prior strikes.
For convenience we designate four distinct phases of this history. Phase I is the period from the initial arrest to the initial sentencing, which in this case continued from February 11, 1995, until July 9, 1996. Phase II is the period from the initial sentencing to the reversal (July 9, 1996, through July 9, 1999). Phase III is the period from the reversal to the second sentencing (July 9, 1999, until Aug. 19, 1999), and phase IV is the period after the second and final sentencing (after Aug. 19, 1999). The parties do not dispute that petitioner should accrue credits as a presentence inmate for phases I and III (see § 4019), and they likewise agree that petitioner should accrue credits as a postsentence second striker for phase IV. The dispute lies in the characterization of the three-year phase II status.
The trial court resentenced petitioner on August 19, 1999, after she had pleaded guilty. The court recalculated the entire period prior to August 19, 1999 (phases I, II and III), as presentence time, granting petitioner conduct credit for 50 percent of her actual custody time. Under this theory, petitioner’s phase II credit accrual (up to 50 percent pursuant to § 4019, subd. (f)) is based on her phase III status as a presentence inmate.
The trial court again resentenced petitioner on April 17, 2001.2 The court granted petitioner credit under the presentence section 4019 formula for phases I and III, and granted petitioner conduct credit in accordance with the limitations expressed in sections 667, subdivision (c)(5) and 1170.12, subdivision (a)(5) for phases II and IV. This sentencing based petitioner’s phase II status on her ultimate phase IV status as a recidivist with one prior strike.
The Court of Appeal endorsed the 1999 sentencing position, reasoning that the reversal of petitioner’s conviction meant the initial conviction ceased to exist. Accordingly, petitioner could not have bee°n a postsentence inmate during phase II because there was no valid conviction and sentence in existence. The Attorney General petitioned for review, implicitly asserting the April 17, 2001, resentencing was correct. We granted review, and now reverse the decision of the Court of Appeal.
*33II. Analysis
A. Section 2900.1
Our analysis begins with section 2900.1: “Where a [petitioner] has served any portion of his sentence under a commitment based upon a judgment which judgment is subsequently declared invalid or which is modified during the term of imprisonment, such time shall be credited upon any subsequent sentence he may receive upon a new commitment for the same criminal act or acts.” No case interpreting section 2900.1 has squarely addressed whether the “time” that “shall be credited” under the statute is presentence or postsentence.
In In re James (1952) 38 Cal.2d 302 [240 P.2d 596] (James), the petitioner pleaded guilty in December 1944 to first degree murder and was sentenced to life imprisonment. (Id. at p. 308.) In January 1952, we determined the plea had been invalid, and reversed the conviction. (Id. at p. 313.) We explained that if the People obtained a conviction for manslaughter on retrial (having apparently agreed not to retry James for murder), he would be entitled to credit for the more than seven years of actual confinement. Furthermore, because we implicitly deemed this confinement postsentence, rather than characterizing it as presentence, we noted James would be eligible for conduct credit pursuant to the postsentence credit statute. (Id. at p. 314.)3 We also declined to compute James’s prereversal credits in accordance with his prereversal status as a life prisoner, which would have rendered him ineligible to accrue credits against a fixed term. (See In re Bentley (1974) 43 Cal.App.3d 988, 995 [118 Cal.Rptr. 452].)
Since James did not expressly analyze the question, its result does not control our decision. Nonetheless, the literal terms of the relevant statutes appear in accord with the result in James. Section 4019, subdivision (a)(4), which petitioner contends should apply, applies to someone “confined in a county jail . . . following arrest and prior to the imposition of sentence for a felony conviction.” By contrast, section 2933, subdivision (a), applies to “persons convicted of a crime and sentenced to state prison.” Phase II occurred after petitioner was convicted of a crime, and that confinement was *34served in state prison, not a local jail. Nevertheless, petitioner offers several arguments contending the time she served during phase II should be characterized as presentence pursuant to section 4019.4 Although petitioner’s arguments have some superficial appeal, we find them ultimately unpersuasive.
B. Postsentence Status Is Not a “Credit Disability”
Since James, the Legislature has instituted determinate sentencing and created a complex array of presentence and postsentence credit schemes to serve various functions. As we observed in Buckhalter, this complexity “ ‘ “is likely to produce some incongruous results and arguable unfairness when compared to a theoretical state of perfect and equal justice. [Because] there is no simple or universal formula to solve all presentence credit issues, our aim [must be] to provide ... a construction [of the statutory scheme] which is faithful to its language, which produces fair and reasonable results in a majority of cases, and which can be readily understood and applied by trial courts.” ’ ” (Buckhalter, supra, 26 Cal.4th at p. 29, quoting People v. Bruner (1995) 9 Cal.4th 1178, 1195 [40 Cal.Rptr.2d 534, 892 P.2d 1277], quoting In re Joyner (1989) 48 Cal.3d 487, 495 [256 Cal.Rptr. 785, 769 P.2d 967].)
Petitioner contends she should not be subject to what she describes as the “credit disability” of postsentence status in light of the subsequent reversal. Petitioner’s argument focuses on the disadvantage (through reduced conduct credit) that she suffers by having her phase II time deemed postsentence time. It is not self-evident, however, that postsentence status is an inherent disability. Whether a petitioner fares better “presentence” or “post-sentence” will vary, depending on the nature of the commitment offense and the petitioner’s history. A nonviolent offender may receive a credit up to 50 percent of her actual presentence confinement. (§ 4019.) If she has no prior strikes, she may earn 100 percent credit postsentence (one day of conduct credit for each day actually served) (§ 2933, subd. (a)), whereas a recidivist with a prior strike may earn postsentence credits only up to 20 percent of the total prison sentence (§§ 667, subd. (c)(5), 1170.12, subd. (a)(5)), and an offender with two prior strikes is denied any postsentence conduct credit. (In re Cervera (2001) 24 Cal.4th 1073, 1076 [103 Cal.Rptr.2d 762, 16 P.3d *35176].)5 Thus, the negative consequences of postsentence status are linked to petitioner’s recidivist status.6 If petitioner had no prior strikes, she would fare better with postsentence status and would probably object if her reversal prompted a recharacterization of her prison custody as presentence time, limiting her conduct credit to 50 percent of her actual custody time.
As we observed in Buckhalter, “the pre- and postsentence credit systems serve disparate goals and target persons who are not similarly situated. The presentence credit scheme, section 4019, focuses primarily on encouraging minimal cooperation and good behavior by persons temporarily detained in local custody before they are convicted, sentenced, and committed on felony charges. By contrast, the worktime credit scheme for persons serving prison terms emphasizes penological considerations, including the extent to which certain classes of prisoners, but not others, deserve or might benefit from incentives to shorten their terms through participation in rehabilitative work, education, and training programs . . . .” (Buckhalter, supra, 26 Cal.4th at p. 36.) These considerations have shaped the distribution of postsentence credit: The Legislature has determined that offenders with no prior strikes are the most amenable to rehabilitation, and they thus earn the most credits. Offenders with one prior strike receive reduced postsentence credit, and those with two prior strikes, considered the least amenable to rehabilitation, receive none at all.
The Legislature has thus enacted the current scheme to ensure a petitioner’s credit accrual is commensurate with her status. This may produce occasional incongruities. For example, granting more credits to postsentence inmates may inadvertently reward offenders who post bail or plead guilty quickly (see, e.g., In re Cleaver (1984) 158 Cal.App.3d 770, 773-775 [204 Cal.Rptr. 835]), whereas favoring presentence inmates may similarly reward offenders who do not post bail or who go to trial (People v. Applin (1995) 40 Cal.App.4th 404, 408, 411 [46 Cal.Rptr.2d 862]). These occasional disparities have not been fatal.
C. Petitioner Is Not Entitled to the Same Credits as an Inmate Awaiting Trial
Petitioner contends that during phase II she was similarly situated to a petitioner who had not yet gone to trial. She asserts that she should be no *36worse off for having suffered a later reversed conviction than would be a hypothetical petitioner who simply had her trial delayed for a comparable period of time. She compares herself with pretrial detainees, and argues she should receive the same two-for-four conduct credits.
A pretrial detainee is not similarly situated to a state prison inmate. (Buckhalter, supra, 26 Cal.4th at p. 36; People v. Caddick (1984) 160 Cal.App.3d 46, 53 [206 Cal.Rptr. 454].) “Pretrial detainee-felons are presumptively innocent and therefore not in need of rehabilitation; prison inmates are conclusively guilty and presumptively in need of rehabilitation. In many cases, the pretrial detainee may make bail at any time, thereby interfering with any continual work or education program. . . . Pretrial detainees have court appearances; they consult with their attorneys and other experts their cases may require. This makes continual work rehabilitation or education [programming] impractical; obviously such interruptions are not a concern for prison inmates. Moreover, the Legislature has not declared its intent to achieve self-sufficiency in the county jails. While the foregoing distinctions [do not exhaust] the differences between the two classes, they are sufficient to demonstrate that the classes are not similarly situated.” (Caddick, at p. 53, fn. omitted.) Accordingly, “the terms prison inmates and pretrial detainees are more than labels which define the difference between persons who have been convicted of a felony and sentenced and those who have simply been charged with [a] felony.” (Ibid., fn. 3.) These contrasting functions of confinement warrant the treatment of petitioner’s phase II custody as postsentence.
Furthermore, petitioner’s proposal to recharacterize her phase II confinement as presentence time would arguably create an equal protection problem even worse than that to which petitioner objects. Suppose a jury convicts two defendants, each of whom has a prior strike, of the same first degree burglary (§§ 459, 460), and imposes the upper base term of six years (§ 461) on each. If there are no errors with A’s trial, his first four years of postsentence custody will yield A one year of conduct credit. (§§ 667, subd. (c)(5), 1170.12, subd. (a)(5).) By contrast, if, after the same four-year period, a reviewing court finds the trial court improperly denied B’s motion for self-representation, petitioner’s theory would grant B two years of credit for this time (§ 4019), even though the exact same evidence was presented against each codefendant.7
D. Petitioner’s Credits Are Not Based on an Incorrect Judgment
Finally, petitioner contends her credit accrual should not be based on an incorrect initial judgment, and should be corrected to apply retroactively to *37her phase II confinement. But her subsequent plea confirms that her initial conviction, although procedurally invalid, was not without legal basis. On the contrary, she pleaded guilty to the charged offense, and thus does not warrant disparate treatment from a petitioner who initially received an error-free judgment.
Petitioner objects to characterizing her phase II custody as postsentence “retroactively,” based on her subsequent guilty, plea. But her entire claim depends on our willingness to recharacterize her phase II custody retroactively, i.e., after her reversal. We agree that petitioner’s phase II time may be recharacterized. Our recharacterization analysis, however, does not cease with her reversal, but also includes her eventual reconviction.
Petitioner derives some benefit, however, from our recharacterization. She was sentenced initially as a third striker, which would have rendered her ineligible to earn any postsentence conduct credits, as in James, supra, 38 Cal.2d 302. She ultimately pleaded guilty as a second striker, eligible to earn postsentence conduct credit up to 20 percent of the total prison sentence. Because we follow petitioner’s ultimate phase IV status, we conclude that during her phase II confinement, she was entitled to earn credits under the 20 percent formula described above.8
Conclusion
It is petitioner’s ultimate phase IV status as a convicted second striker, not her unresolved phase III status as presentence petitioner, nor her initial phase II status as a convicted third striker, that must control the determination of her phase II credits. Petitioner pleaded guilty as having one prior strike, and she thus is entitled, for her phase II confinement, to a maximum conduct credit of 20 percent of her total prison sentence.
We therefore reverse the judgment of the Court of Appeal. The matter is remanded to the Court of Appeal with instructions to direct the trial court to sentence petitioner in accordance with the April 17, 2001, sentencing. The trial court should clarify the date on which petitioner’s custody commenced. Calculation of the actual days of conduct credit earned by petitioner during her phase II and phase IV confinements should be left to prison authorities. (See Buckhalter, supra, 26 Cal.4th at pp. 30-31, 40-41.)
George, C. J., Baxter, J., Chin, J., and Moreno, J., concurred.
All further statutory references are to the Penal Code.
The April 17, 2001, abstract of judgment states petitioner’s custody commenced on May 19, 1995, although it appears her custody actually commenced on February 11, 1995. The abstract of judgment’s custody calculations, however, reflected the number of days served (515) since February 11.
The extant credit scheme authorized conduct credits of up to five days per month for county jail inmates (Stats. 1941, ch. 106, § 15, pp. 1122-1123 [adding § 4019]), whereas state prisoners could earn two months of conduct credit per year for the first two years of confinement, four months per year for the next two years, and five months per year thereafter. (Stats. 1941, ch. 106, § 15, pp. 1105-1106 [adding § 2920].) These statutes were repealed and replaced with more generous conduct credit accrual provisions with the advent of determinate sentencing. (See Stats. 1976, ch. 286, §§ 3, 4, pp. 595-596; Stats. 1977, ch. 165, §§ 36, 37, p. 661.)
The parties’ dispute concerns phase II time only; we therefore need not express any opinion as to the proper characterization of phase III time.
Some individuals accrue conduct credit at the same rate for both their presentence and postsentence custody. Violent felons receive the same 15 percent credit (§ 2933.1, subd. (a)), and convicted murderers receive no credit at all (§ 2933.2, subd. (a)).
Although a pretrial detainee is presumed innocent until proven guilty, this does not, as shown above, necessarily entitle a petitioner to more favorable credit provisions. In any event, petitioner was convicted on December 11, 1995, and thus eight of the 17 months of presentence custody elapsed after her conviction.
We further note that our hypothetical is more likely to create great disparities than petitioner’s. Few defendants will spend several years in pretrial custody. Contrariwise, prisoners often obtain reversal years after their initial conviction.
Neither the People nor the petitioner asserts the initial sentencing of petitioner as a “third striker” should control her credit accrual for that period. If petitioner were denied all conduct credit due to the initial conviction, it could create an equal protection problem against a hypothetical petitioner who initially pleaded guilty to having one prior strike and thus received 20 percent conduct credits for her entire confinement.