Darren Bottinelli and seven other federal prisoners ("Petitioners") appeal from the district court's denial of their joint petition for a writ of habeas corpus. They argue that the recently enacted First Step Act's amendment to the good time credit provision requires the Bureau of Prisons ("BOP") to re-calculate their sentences immediately, which would accelerate their dates for release or transfer to prerelease custody. We have jurisdiction under 28 U.S.C. § 1291, and we affirm.
I. BACKGROUND
A federal prisoner who is serving more than a one-year term of imprisonment may earn good time credit toward his or her sentence so long as the prisoner "display[s] exemplary compliance with institutional disciplinary regulations." 18 U.S.C. § 3624(b)(1). Section 3624(b) controls how the BOP calculates good time credit. Until recently, although the statute provided that prisoners could earn "up to 54 days" each year for exemplary compliance, the BOP used a calculation that allowed a maximum of only 47 days. See Barber v. Thomas , 560 U.S. 474, 476-79, 130 S.Ct. 2499, 177 L.Ed.2d 1 (2010) (discussing § 3624(b)(1) ). Courts, including ours, upheld this 47-day calculation. See id . at 492, 130 S.Ct. 2499 ; Pacheco-Camacho v. Hood , 272 F.3d 1266, 1271 (9th Cir. 2001).
On December 21, 2018, the First Step Act of 2018, Pub. L. No. 115-391, 132 Stat. 5194, was enacted. The Act implemented a number of prison and sentencing reforms. We limit our focus to subsection 102(b), which made two amendments relevant to this appeal.
First, paragraph 102(b)(1) amends § 3624(b) - the good time credit provision - to require the BOP to permit up to 54 days per year. § 102(b), 132 Stat. at 5210. Second, paragraph 102(b)(1) amends § 3624 by adding subsection (g), which is relevant to the Act's creation of an earned time credit system.1 Id. at 5210-13. The Act requires that, within 210 days of its enactment, the Attorney General establish a "risk and needs assessment system" to, broadly speaking, review each prisoner's recidivism risk level, award earned time *1198credit as an incentive for participation in recidivism reduction programming, and "determine when a prisoner is ready to transfer into prerelease custody or supervised release in accordance with section 3624." § 101(a), 132 Stat. at 5196-97. Section 3624(g) details the criteria for when a prisoner becomes eligible, considering earned time credit, for transfer to prerelease custody or supervised release. § 102(b), 132 Stat. at 5210-13.
Subsection 102(b) also includes an effective-date provision in paragraph 102(b)(2), and a retroactivity provision in paragraph 102(b)(3).
SEC. 102. IMPLEMENTATION OF SYSTEM AND RECOMMENDATIONS BY BUREAU OF PRISONS.
(b) PRERELEASE CUSTODY.- ...
(2) EFFECTIVE DATE.-The amendments made by this subsection shall take effect beginning on the date that the Attorney General completes and releases the risk and needs assessment system under subchapter D of chapter 229 of title 18, United States Code, as added by section 101(a) of this Act.
(3) APPLICABILITY.-The amendments made by this subsection shall apply with respect to offenses committed before, on, or after the date of enactment of this Act, except that such amendments shall not apply with respect to offenses committed before November 1, 1987.
Id . at 5208, 5210, 5213.
The parties agree that the Act now provides federal prisoners the possibility of seven additional days of good time credit per year. They disagree, however, as to when that amendment takes effect. Petitioners argue that the amendment took effect upon the Act's enactment on December 21, 2018, and, therefore, that they are entitled to the immediate recalculation of their good time credit. But the BOP contends that the amendment does not take effect until July 19, 2019 - the date by which the Attorney General must establish "the risk and needs assessment system."
The district court agreed with the BOP. It held that the Act's text clearly links the good time credit amendment's effective date to the creation of "the risk and needs assessment system." Accordingly, the court explained that it "cannot, as Petitioners invite [it] to do, ignore Section 102(b)(2)'s express and unambiguous text and conclude that, despite what it clearly said, Congress really intended the 'good time fix' to be effective immediately."2
II. DISCUSSION
We review de novo the denial of a 28 U.S.C. § 2241 petition, *1199Stephens v. Herrera , 464 F.3d 895, 897 (9th Cir. 2006), and questions of statutory interpretation, Chemehuevi Indian Tribe v. Newsom , 919 F.3d 1148, 1151 (9th Cir. 2019).
A. Statutory Interpretation
The single issue before us is when the Act's good time credit amendment takes effect. As a general rule, a statute takes effect upon its enactment unless Congress clearly provides otherwise. See Gozlon-Peretz v. United States , 498 U.S. 395, 404, 111 S.Ct. 840, 112 L.Ed.2d 919 (1991). "When interpreting a statute, we are guided by the fundamental canons of statutory construction and begin with the statutory text." United States v. Neal , 776 F.3d 645, 652 (9th Cir. 2015). Whether the statutory text has a "plain and unambiguous meaning" depends on "the language itself, the specific context in which that language is used, and the broader context of the statute as a whole." United States v. Youssef , 547 F.3d 1090, 1093 (9th Cir. 2008) (per curiam) (quoting Robinson v. Shell Oil Co. , 519 U.S. 337, 340-41, 117 S.Ct. 843, 136 L.Ed.2d 808 (1997) ).
We agree with the district court's reading of the Act. Congress provided "clear direction" in paragraph 102(b)(2) to delay the implementation of the good time credit amendment until the Attorney General establishes the "risk and needs assessment system." Gozlon-Peretz , 498 U.S. at 404, 111 S.Ct. 840. Like the district court, we refuse to overlook the Act's plain text and therefore affirm the denial of the joint habeas petition.
Our analysis can begin and end with paragraph 102(b)(2)'s text. First, paragraph 102(b)(2)'s title - " EFFECTIVE DATE " - leaves no doubt about its purpose. § 102(b), 132 Stat. at 5213. Second, paragraph 102(b)(2)'s text unambiguously applies to the entire subsection. Its subject - "[t]he amendments made by this subsection" - refers to "amendments" in the plural without any qualifications. Id. Subsection 102(b) makes only two amendments: the good time credit amendment and the addition of § 3624(g). Id. at 5210-13. Therefore, the only logical conclusion is that paragraph 102(b)(2) must apply to both amendments made in paragraph 102(b)(1). Third, paragraph 102(b)(2)'s text is straightforward as to when these amendments become operative: "The amendments made by this subsection shall take effect beginning on the date that the Attorney General completes and releases the risk and needs assessment system ... as added by section 101(a) of this Act." Id. at 5213 (emphasis added). In addition to the mandatory language - "shall take effect" - paragraph 102(b)(2) plainly ties the amendments' effectivity to the date that the "risk and needs assessment system" is established. Although no specific date is provided, the text refers directly to "section 101(a) of this Act," which requires the Attorney General to establish the "risk and needs assessment system" within 210 days of enactment, thus no later than July 19, 2019. § 101(a), 132 Stat. at 5196.
Conceding that paragraph 102(b)(2)'s text is difficult to overcome, Petitioners argue that "the full statutory context" requires reading paragraph 102(b)(2) so that it does not apply to the good time credit amendment. They contend that "this subsection" in paragraph 102(b)(2) refers only to paragraph 102(b)(1)'s addition of § 3624(g) because just that amendment is related to the newly created "risk and needs assessment system."
We reject this narrow reading of "this subsection." Interpreting paragraph 102(b)(2) to apply only to the addition of § 3624(g) is unnatural. This proposed interpretation also asks us to ignore how "this subsection" is used elsewhere. See Barber , 560 U.S. at 483-84, 130 S.Ct. 2499 (recognizing a "presumption that a given *1200term is used to mean the same thing throughout a statute" (citation omitted)). For instance, paragraph 102(b)(3) uses "this subsection" to give both of paragraph 102(b)(1)'s amendments retroactive effect:
SEC. 102. IMPLEMENTATION OF SYSTEM AND RECOMMENDATIONS BY BUREAU OF PRISONS.
(b) PRERELEASE CUSTODY.- ...
(3) APPLICABILITY.-The amendments made by this subsection shall apply with respect to offenses committed before, on, or after the date of enactment of this Act, except that such amendments shall not apply with respect to offenses committed before November 1, 1987.
§ 102(b), 132 Stat. at 5208, 5210, 5213. As in paragraph 102(b)(2), the subject of paragraph 102(b)(3) - "[t]he amendments made by this subsection" - refers to more than a single amendment and contains no qualifier to limit its application to only the addition of § 3624(g). Id. Notably, Petitioners do not advocate for a narrow interpretation of "this subsection" in paragraph 102(b)(3) because that would produce a disadvantageous result for them.
Still in search of a viable argument for narrowly construing paragraph 102(b)(2)'s text, Petitioners contend that we should not interpret the Act as written because it is illogical to link the good time credit amendment's effective date to the "risk and needs assessment system." At first glance, Petitioners are correct that it is not immediately intuitive why Congress linked the two. However, even though the good time credit calculation and the "risk and needs assessment system" are not dependent on each other, they are generally related. Specifically, both affect prerelease custody under 18 U.S.C. § 3624(c). The good time credit calculation affects a prisoner's projected release date, which affects when a prisoner can be transferred to prerelease custody. See 18 U.S.C. § 3624(c)(1) (requiring that the BOP, "to the extent practicable, ensure that a prisoner serving a term of imprisonment spends a portion of the final months of that term" in prerelease custody). And, under the new "risk and needs assessment system," receiving earned time credit can potentially shorten prerelease custody or supervised release. See § 101(a), 132 Stat. at 5198. Because good time credit does not operate entirely separately from the "risk and needs assessment system," Congress rationally placed the good time credit amendment in subsection 102(b), titled " PRERELEASE CUSTODY ." § 102(b), 132 Stat. at 5210; see also Almendarez-Torres v. United States , 523 U.S. 224, 234, 118 S.Ct. 1219, 140 L.Ed.2d 350 (1998) (" '[T]he title of a statute and the heading of a section' are 'tools available for the resolution of a doubt' about the meaning of a statute." (citation omitted)).
Lastly, we reject Petitioners' assertion that applying the Act's text as written would produce absurd results, as the good time credit amendment requires more than a simple calculation tweak. Releasing a prisoner comes with a host of obligations, including the provision of clothing, a stipend, and transportation. See 18 U.S.C. § 3624(d). Because the good time credit amendment affects a large swath of the federal incarcerated population, there is nothing absurd about giving the BOP time to prepare for this increased administrative burden and the ensuing number of releases and prerelease transfers.
B. Alleged Drafting Error
Likely understanding the weakness in their proposed statutory interpretation, Petitioners move beyond textual arguments. They argue that even if paragraph 102(b)(2) delays the good time credit amendment, this is a "drafting oversight" for us to fix. We find nothing to suggest that the delay was such an oversight. Moreover, Petitioners' reliance here on *1201Gozlon-Peretz is inapt. There, the United States Supreme Court clarified when a statute, which was without an effective-date provision, became operative. Gozlon-Peretz , 498 U.S. at 404, 111 S.Ct. 840 ("[W]e note that § 1002, like many other congressional enactments, contains no provision for its effective date. Nor is there an effective date specified for the [relevant Act] as a whole."). Because it could not "say that Congress gave a clear direction to delay the effective date," the Court applied the general principle that a statute takes effect upon enactment. Id. at 407, 111 S.Ct. 840.
The Act before us is different. Subsection 102(b) is not silent as to when its amendments take effect. Because Congress gave us "clear direction to delay the effective date," we begin our analysis at a different juncture, and Gozlon-Peretz has no application. Id. ; see also Mobil Oil Corp. v. Higginbotham , 436 U.S. 618, 625, 98 S.Ct. 2010, 56 L.Ed.2d 581 (1978) ("There is a basic difference between filling a gap left by Congress' silence and rewriting rules that Congress has affirmatively and specifically enacted."). Generally, we hesitate to presume a statutory drafting error, and nothing here requires us to deviate from that principle. See Lamie v. U.S. Tr. , 540 U.S. 526, 542, 124 S.Ct. 1023, 157 L.Ed.2d 1024 (2004) ("It is beyond our province to rescue Congress from its drafting errors, and to provide for what we might think ... is the preferred result." (citation omitted)).
C. Constitutional Claim
Finally, we also reject Petitioners' contention that the delay violates the Constitution. Petitioners argue that reading paragraph 102(b)(2) to stall the good time credit amendment violates the Equal Protection Clause because federal prisoners with projected release dates between December 21, 2018, and July 19, 2019, are irrationally deprived of the amendment's benefit and thus overserve their sentences. Because Petitioners do not allege that this classification implicates a suspect class or infringes on a fundamental right, we apply only rational basis review.
Petitioners have failed to satisfy their burden of "disprov[ing] the rationality of the relationship between the classification and the purpose." United States v. Navarro , 800 F.3d 1104, 1113 (9th Cir. 2015) (citation omitted); see also United States v. Padilla-Diaz , 862 F.3d 856, 862 (9th Cir. 2017) ("[A] classification is valid 'if there is any reasonably conceivable state of facts that could provide a rational basis for the classification.' " (citation omitted)). As already discussed, Congress understandably gave the BOP time to implement the good time credit amendment rather than making it immediately effective. This respite eased the BOP's administrative burden and bolstered its ability to prepare for the impending releases and prerelease transfers. Questioning whether the Act provided the BOP too many days to prepare does not overcome the deference we owe to Congress. See McGinnis v. Royster , 410 U.S. 263, 276, 93 S.Ct. 1055, 35 L.Ed.2d 282 (1973) (stating that "legislative solutions must be respected if the 'distinctions drawn have some basis in practical experience' " (citation omitted)).
Significantly, we recently rejected a similar constitutional argument. In Navarro , the defendant challenged the United States Sentencing Commission's delayed implementation of a sentence-reduction amendment because he did not benefit unless that amendment took immediate effect. 800 F.3d at 1107-08. Denying the equal protection challenge, we recognized the "legitimate government interests" in delay because of the large number of potentially eligible prisoners and the agency's need to prepare. Id. at 1113-14. Here, Congress similarly had appropriate reasons to delay the amendment's implementation, *1202and thus the classification easily survives rational basis review.3 See Padilla-Diaz , 862 F.3d at 862 (explaining that a statute "will sometimes produce unequal and arguably unfair results" yet survive rational basis review).
III. CONCLUSION
We hold that the Act's good time credit amendment did not take immediate effect upon enactment but will become effective with the establishment of the "risk and needs assessment system" on July 19, 2019.
AFFIRMED .