— Following this court’s decision in In re Mota, 114 Wn.2d 465, 472, 788 P.2d 538 (1990), holding the Department of Corrections’ (DOC) denial of earned early release credit to presentence detainees in county jails violated equal protection, the Legislature amended the sentencing statute to authorize the county jails to establish earned early release credit policies. RCW 9.94A. 150(1). Now we are asked to review the constitutionality of those county jail policies: presentence detainees receive earned early release credit, but less than that available under DOC policy. We hold county jail policies setting a maximum earned early release credit lower than the DOC policy are statutorily and constitutionally permissible.
As mandated by Mota, the Legislature requires all prisoners, including those in presentence' detention, be *59eligible for earned early release credit whether held in a DOC facility or a county jail:
Except as otherwise provided for in subsection (2) of this section, the term of the sentence of an offender committed to a correctional facility operated by the department, may be reduced by earned early release time in accordance with procedures that shall be developed and promulgated by the correctional agency having jurisdiction in which the offender is confined. The earned early release time shall be for good behavior and good performance, as determined by the correctional agency having jurisdiction. The correctional agency shall not credit the offender with earned early release credits in advance of the offender actually earning the credits. Any program established pursuant to this section shall allow an offender to earn early release credits for presentence incarceration. If an offender is transferred from a county jail to the department of corrections, the county jail facility shall certify to the department the amount of time spent in custody at the facility and the amount of earned early release time. In the case of an offender convicted of a serious violent offense or a sex offense that is a class A felony committed on or after July 1, 1990, the aggregate earned early release time may not exceed fifteen percent of the sentence. In no other case shall the aggregate earned early release time exceed one-third of the total sentence!.]
RCW 9.94A. 150(1).
Pursuant to RCW 9.94A.150(1), the DOC and the county jails have developed the independent credit policies at issue here. The DOC’s earned early release credit system, also known as "good-time,” consists of two types of credit: good-conduct time and earned time. A prisoner in a state facility earns "good-conduct time” by good behavior at ten days per thirty days served; generally, the prisoner is presumed to have earned the full good-conduct time available unless the facility specifically deducts for misbehavior. In re Cromeenes, 72 Wn. App. 353, 355, 864 P.2d 423 (1993). The DOC will award an additional five days per thirty days served in "earned time” for participation in work, academic, or treatment programs. Cromeenes, 72 Wn. App. *60at 355; RCW 9.95.070. The DOC then combines the total good-conduct and earned time, up to one-half of the days served, to determine the final sentence reduction, up to the statutory maximum of one-third of the imposed sentence. In re Williams, 121 Wn.2d 655, 659, 853 P.2d 444 (1993); Cromeenes, 72 Wn. App. at 355; RCW 9.94A.150(1).
A prisoner denied or unable to pay bail will spend presentence detention in a county jail and then transfer to a DOC facility upon sentencing. In that case, the DOC will reduce the prisoner’s sentence not only for earned early release credit accumulated at the DOC facility, but also for day-for-day credit and earned early release credit earned in presentence detention. Williams, 121 Wn.2d at 658-59; State v. Phelan, 100 Wn.2d 508, 517, 671 P.2d 1212 (1983). Defendant James N. Fogle spent 102 days of presentence detention in the Pierce County Jail on a sixty-month sentence, earning fifteen days earned early release credit. Defendant Donald D. MacFarlane spent 144 presentence days in the Clark County Jail on concurrent sentences of seventy-two and eight months, earning twenty-one earned early release credit days.
Like the DOC, the Clark and Pierce County Jails use tiered credit systems. The jails’ version of good-conduct time is known as "good time,” the base award available to all general population prisoners at a maximum fifteen percent of the imposed sentence; the jail will deduct any misconduct from that credit. Comparable to the DOC’s earned time, the jails’ policies permit prisoners to receive additional credit through work programs to achieve a maximum earned early release credit of thirty percent of the imposed sentence at Clark County Jail and twenty percent or thirty percent at Pierce County Jail. Both Fogle and MacFarlane received the maximum fifteen percent credit as general population prisoners.
Defendants filed separate personal restraint petitions in the court of appeals attacking the county jails’ earned early release policies. Denying an equal protection chai*61lenge, the court of appeals granted Fogle’s petition in part on the basis that the Pierce County Jail inaccurately calculated his credit and ordered readjustment to seventeen days. This court granted Fogle direct discretionary review and certified MacFarlane’s petition, consolidating the cases on review.
Defendants focus on two features of the county jail policies: the denial of the full statutory maximum and the exclusion of most prisoners from the higher credit work programs. Defendants claim the county jails have exceeded their statutory authority by formulating these policies. At the heart of this case, Defendants raise an equal protection challenge against the disparate treatment of presentence detainees. In addition, Defendants claim the policies violate double jeopardy and due process.
I
Statutory Authority for Earned Early Release Credit
Defendants argue RCW 9.94A.150(1) mandates all earned early release credit programs allow every prisoner, except those explicitly excluded as violent or sex offenders, to earn the full statutory maximum credit. Defendants’ argument runs contrary to this court’s decision in In re Williams, 121 Wn.2d 665, and the Legislature’s intent for independent, plenary county jail authority over earned early release policies. See Williams, 121 Wn.2d at 666.
As here, Williams considered an equal protection challenge to a county jail’s certification of less than the statutory maximum for a presentence detainee. Without reaching the equal protection issue, Williams rested on the issue of statutory authority to hold prisoners do not have "a right ... to challenge their [jail] certifications merely because they were not awarded the statutorily permitted maximum amount of good time.” Williams, 121 Wn.2d at 666. Nor does the statute create a right to a specific method of calculation.
*62We also reject Defendants’ interpretation of RCW 9.94A.150(1) as requiring that jails permit all prisoners access to work programs. The clause describing "good behavior and good performance” merely indicates the factors for awarding credit within the provision’s overall grant of authority to facilities to create independent policies. Williams, 121 Wn.2d at 661. We hold the Clark and Pierce County Jails’ policies are consistent with the grant of authority in RCW 9.94A. 150(1).
II
Equal Protection Analysis
"Equal protection requires that persons similarly situated receive like treatment.” Mota, 114 Wn.2d at 473. Defendants assert the Clark and Pierce County Jails’ earned early release policies violate prisoners’ equal protection rights under the state and federal constitutions by awarding most prisoners, particularly presentence detainees, only fifteen percent credit. Defendants claim the jails’ policies result in longer sentences for presentence detainees than for prisoners spending their entire sentences in a DOC facility.
A. Level of Scrutiny
Equal protection analysis generally applies the rational basis test: whether the policy is rationally related to a legitimate state purpose. Mota, 114 Wn.2d at 475. In Mota, however, this court utilized intermediate scrutiny, holding the state failed to assert a substantial interest in denying presentence detainees any earned early release credit. Mota, 114 Wn.2d at 473. Defendants urge the court to adopt an intermediate scrutiny standard here.
Intermediate scrutiny is appropriate only where the denial of a liberty interest is due to a semisuspect classification. Mota, 114 Wn.2d at 474. The Mota court reasoned the DOC’s failure to award good-time credit under the former RCW 9.94A.150(1) to a presentence *63detainee financially unable to post bail represented discrimination on the basis of wealth. Mota, 114 Wn.2d at 474. Following Mota, the court of appeals has applied rational basis to an equal protection challenge to disparate earned early release policies where the presentence detainee was held without bail: indigency did not serve as the basis for the disparate treatment. Cromeenes, 72 Wn. App. at 357.
Without disturbing Mota, we note failure to pay set bail does not necessarily represent a wealth-based classification to merit semisuspect status. The determination of bail may depend on many factors beyond wealth, such as perceived dangerousness and likelihood of flight. See CrR 3.2(b). Moreover, a prisoner may elect not to pay bail for reasons other than financial condition.
In the present case there is some dispute whether Defendants were eligible for bail. In addition, both Defendants were held on separate fugitive warrants, likely to render payment of bail on the primary charge futile. Nonetheless, we will construe the facts in the light most favorable to Defendants and proceed on an intermediate scrutiny standard.
B. Intermediate Scrutiny Analysis
Defendants claim no substantial, or even legitimate state interest, is served by the county jails’ tiered credit policies awarding only fifteen percent credit to most prisoners. Although thirty percent credit is available to county jail prisoners, Defendants submit most inmates, especially presentence detainees, do not achieve work program status. Applying intermediate scrutiny, we hold the state’s substantial interest in maintaining prisoner discipline, particularly by preventing flight from prosecution and preserving local control over jails, justifies disparate treatment to overcome Defendants’ equal protection challenge.
We first remark, contrary to Defendants’ position, the result of the fifteen percent award for general population *64prisoners does not result in longer sentences for presentence detainees. Earned early release credit has no effect on sentencing. The lower award does result in more days served. Had Defendants served their presentence days in a DOC facility they could have potentially earned more credit and ultimately a larger sentence reduction. On good-conduct time alone, Fogle would have been eligible for thirty-four days, MacFarlane for forty-eight days. With earned time Fogle might have reached fifty-one days and MacFarlane seventy-two days.
Discipline and punishment are the primary purposes of the sentencing statute. Williams, 121 Wn.2d at 662; Mota, 114 Wn.2d at 475-77. Limiting access to work programs serves the substantial purposes of punishment and discipline by basing eligibility on the severity of the crime and the prisoner’s behavior. See Williams, 121 Wn.2d at 662. In regard to program participation, presentence detainees are not similarly situated to convicted prisoners in county jail or a DOC facility: the risk of flight from prosecution makes presentence detainees particularly unsuitable for work program assignment. Cromeenes, 72 Wn. App. at 358. As we explained in Williams:
[T]he very fact of bail and presentence incarceration raises the possibility of disparate treatment based upon wealth. In general, however, the needs of the justice system in assuring the presence of defendants at trial are deemed sufficient to validate such a system. . . .
(Citations omitted.) Williams, 121 Wn.2d at 665. The tiered system also advances discipline by providing an incentive for county inmates to improve behavior and earn work program participation. Cromeenes, 72 Wn. App. at 357.
The Legislature authorized independent county jail credit policies in RCW 9.94A.150 in recognition that the statute’s disciplinary purpose was better promoted through local control. As this court emphasized in Williams:
If a correctional institution is to be able to effectively *65administer its good-time policies in order to further prison discipline, it must actually have control over the award or denial of good time for offenders under its jurisdiction. Good time would be useless in controlling prison discipline in county jails if offenders knew they would be automatically credited with full good time upon their transfer to the Department. . . .
Williams, 121 Wn.2d at 662. The county jails are uniquely situated to determine the best means to control their prisoner populations, particularly presentence detainees presenting special disciplinary concerns not faced by DOC facilities.
Ill
Double Jeopardy
Defendants maintain the jails’ denial of earned early release credit equal to that available from the DOC imposes multiple punishment to violate the prohibition against double jeopardy. As this court noted in State v. Phelan, 100 Wn.2d 508, double jeopardy demands full credit for presentence jail time as well as earned early release time actually earned. Phelan, 100 Wn.2d at 515. At the same time, double jeopardy does not mandate earned early release credit for speculative time the prisoner might have earned under a different credit program. Phelan, 100 Wn.2d at 516. Reaffirming the Phelan analysis, we hold the disparate earned early release policies here do not result in multiple punishment to trigger double jeopardy protection. See Phelan, 100 Wn.2d at 516.
IV
Due Process
Defendants contend the jails violated due process by failing to establish procedures for presentence detainees to earn the statutory maximum earned early release credit. This court has recognized the statutory right to *66earned early release credit creates a limited liberty interest requiring minimal due process. In re Anderson, 112 Wn.2d 546, 548, 772 P.2d 510, cert. denied, 493 U.S. 1004 (1989). However, where due process concerns are implicated in an equal protection challenge, the court will generally rest its decision on an equal protection analysis. Mota, 114 Wn.2d at 474 n.1. Regardless, the jails here established and observed their written earned early release policy consistent with due process requirements.
Durham, C.J., and Smith, Guy, and Talmadge, JJ., concur.