OPINION

TALLMAN, Circuit Judge:

In 2013, the California legislature enacted Assembly Bill 97 (AB 97), a massive reform package designed to streamline public education financing and decentralize education governance. 2013 Cal. Legis. Serv. ch. 47 (A.B. 97). This appeal asks us to consider whether AB 97 abrogated our decisions in Belanger v. Madera Unified School District, 963 F.2d 248 (9th Cir. 1992), and Eaglesmith v. Ward, 73 F.3d 857 (9th Cir. 1995), in which we held that California school districts and county offices of education (COEs) are "arms of the state" entitled to state sovereign immunity. We hold that the passage of AB 97 had no such effect. School districts and COEs in California remain arms of the state and cannot face suit.

I

Defendant Orange County Department of Education (OCDE) hired plaintiff Michael Sato as a Systems Database Architect in August 2014. Within a matter of weeks after Sato started working at OCDE, Sato's supervisors informed him that he would be terminated immediately. OCDE offered no explanation as to why Sato was being terminated, and Sato insisted that he had performed his duties satisfactorily during his brief period of probationary employment with OCDE. Before he was fired, Sato received no oral or written notice of his termination, and he was given no opportunity to be heard at a pre- or post-termination proceeding.

Sato sued OCDE for damages in federal district court, asserting claims under 42 U.S.C. § 1983 and state law. In his complaint, Sato alleged that, under his employment contract with OCDE, Sato could only be fired for cause, even during his initial one-year probationary period.1 Sato claimed that his termination without prior notice or a pre- or post-termination hearing therefore violated his Fourteenth Amendment substantive and procedural due process rights and constituted breach of contract.

OCDE moved to dismiss Sato's complaint under Federal Rules of Civil Procedure 12(b)(1) and 12(b)(6),2 arguing that, pursuant to our decision in Belanger, OCDE is an arm of the state and enjoys Eleventh Amendment immunity from suit.3 Sato responded that "the California Legislature's massive 2013 enactment of Assembly Bill 97 completely replaces the prior statutory basis for, and upends the analysis of, Belanger." According to Sato, after the passage of AB 97, OCDE is no longer an arm of the state for sovereign immunity purposes.

The district court granted OCDE's Rule 12(b)(6) motion with respect to Sato's constitutional claims, but denied OCDE's motion with respect to the breach of contract claim.4 On the issue of sovereign immunity, the district court held that while the passage of AB 97 had perhaps decentralized state control over school funding and governance to some extent, Sato failed to show that AB 97 undermined our reasoning in Belanger. Recognizing that OCDE's sovereign immunity defense applied equally to his federal and state claims, Sato then voluntarily dismissed his state breach of contract claim. After granting Sato's voluntary dismissal motion, the district court entered final judgment in favor of OCDE. We have jurisdiction under 28 U.S.C. § 1291, and we affirm.

II

A

We review the district court's grant of OCDE's Rule 12(b)(6) motion de novo, "accepting as true all well-pleaded allegations of fact in the complaint and construing them in the light most favorable to the plaintiffs." Eason v. Clark Cty. Sch. Dist., 303 F.3d 1137, 1140 (9th Cir. 2002) (quoting Zimmerman v. City of Oakland, 255 F.3d 734, 737 (9th Cir. 2001)). We also review de novo whether a party is entitled to Eleventh Amendment sovereign immunity. Id. "[A]n entity invoking Eleventh Amendment immunity bears the burden of asserting and proving those matters necessary to establish its defense." Del Campo v. Kennedy, 517 F.3d 1070, 1075 (9th Cir. 2008) (internal quotation mark omitted) (quoting In re Lazar, 237 F.3d 967, 974 (9th Cir. 2001)).

Whether the passage of AB 97 abrogates our arm-of-the-state analysis in Belanger presents us with an issue of first impression. While we have applied Belanger and dismissed suits against California school districts since AB 97 was enacted, none of those cases considered AB 97's potential effects on Belanger. See, e.g., Davis v. Folsom Cordova Unified Sch. Dist., 674 Fed.Appx. 699, 701 (9th Cir. 2017); Chadam v. Palo Alto Unified Sch. Dist., 666 Fed.Appx. 615, 618 (9th Cir. 2016); Pierce v. Santa Maria Joint Union High Sch. Dist., 612 Fed.Appx. 897, 898 (9th Cir. 2015); C.W. v. Capistrano Unified Sch. Dist., 784 F.3d 1237, 1247 (9th Cir. 2015); Brynjolfsson v. State Agency L.A. Unified Sch. Dist., 576 Fed.Appx. 697, 698 (9th Cir. 2014). Because none of those cases raised the particular question that confronts us today, we are not bound by them. Cf. Minority Television Project, Inc. v. FCC, 736 F.3d 1192, 1211 (9th Cir. 2013) ("Courts `are not bound by a prior exercise of jurisdiction in a case where it was not questioned and it was passed sub silentio.'" (quoting United States v. L.A. Tucker Truck Lines, Inc., 344 U.S. 33, 38, 73 S.Ct. 67, 97 S.Ct. 54 (1952))).

B

The Eleventh Amendment provides that "[t]he Judicial power of the United States shall not be construed to extend to any suit in law or equity, commenced or prosecuted against one of the United States by Citizens of another State, or by Citizens or Subjects of any Foreign State." U.S. Const. amend. XI. "It is well established that agencies of the state are immune under the Eleventh Amendment from private damages or suits for injunctive relief brought in federal court." Savage v. Glendale Union High Sch., 343 F.3d 1036, 1040 (9th Cir. 2003). State sovereign immunity does not extend to county and municipal governments, unless state law treats them as arms of the state. Id. at 1040-41. The question before us is thus whether, after AB 97, OCDE "is to be treated as an arm of the State partaking of the State's Eleventh Amendment immunity, or instead is to be treated as a municipal corporation or other political subdivision to which the Eleventh Amendment does not extend." See Holz v. Nenana City Pub. Sch. Dist., 347 F.3d 1176, 1180 (9th Cir. 2003) (internal quotation mark omitted) (quoting Mt. Healthy City Sch. Dist. Bd. of Educ. v. Doyle, 429 U.S. 274, 280, 97 S.Ct. 568, 50 L.Ed.2d 471 (1977)).

To determine whether a government entity is an arm of the state, we examine the five factors set forth in Mitchell v. Los Angeles Community College District, 861 F.2d 198 (9th Cir. 1988):

Belanger, 963 F.2d at 250-51 (quoting Mitchell, 861 F.2d at 201). We apply these factors to the question at hand.

AB 97 reformed education funding and governance in California in two key respects. First, it replaced the old mechanism by which the state funded public education, which relied on a combination of "revenue limits" funding and "categorical funding grants," with the Local Control Funding Formula (LCFF). The LCFF streamlined the number of state funding sources and increased K-12 spending, particularly for low-income and English-language-learner (ELL) student populations. Second, AB 97 required that school districts and COEs develop Local Control and Accountability Plans (LCAPs) to measure progress toward state and district goals for student achievement.

Under the Mitchell factors, we conclude that AB 97 did not upend Belanger and Eaglesmith, and we affirm dismissal of Sato's complaint.

1

The first Mitchell factor — whether a money judgment against the government entity would be satisfied out of state funds — is the most important. Eason, 303 F.3d at 1141; see also Savage, 343 F.3d at 1041 ("Because the impetus of the Eleventh Amendment is the prevention of federal-court judgments that must be paid out of a state's treasury, `[t]he vulnerability of the State's purse [is] the most salient factor in Eleventh Amendment determinations.'" (alterations in original) (quoting Hess v. Port Auth. Trans-Hudson Corp., 513 U.S. 30, 48, 115 S.Ct. 394, 130 L.Ed.2d 245 (1994))). In evaluating this factor, "the relevant inquiry is whether [the state] will be legally required to satisfy any monetary judgment obtained against the [d]istrict," not whether state funds will actually be used. Eason, 303 F.3d at 1142; see also Holz, 347 F.3d at 1182 (finding the first Mitchell factor disfavored immunity for the defendant school district, even though the district received 98% of its funding from the state, because Alaska law disclaimed the state's responsibility for school district debts).

In Belanger, we explained that California's unique revenue limits system resulted in the state's legal liability for judgments against school districts. 963 F.2d at 252. Under this system, state law established a maximum per-pupil5 spending amount, equalized across school districts. Id. The state computed the "revenue limit" for each school district by multiplying the statutory per-pupil funding allotment by the number of students in that district. Id. To ensure districts and COEs were funded to the revenue limit, the state subtracted the value of local property taxes from the revenue limit and contributed the remaining amount. Id. (citing Cal. Educ. Code §§ 41600-41610, 42238-42251 (1992)). Administering this funding scheme meant that "state and local revenue [was] commingled in a single fund under state control." Id. And because revenue limits established a maximum, equalized funding level,

Id. In other words, to guarantee that districts and COEs were funded at the statutory revenue limit, the state would be required to backfill any outlay of funds used to satisfy a judgment against a district or COE. Revenue limits thus had the effect of making the state legally liable for any judgment against a school district.

Outside of California, in states that set a minimum, rather than a maximum, per-pupil funding amount, we have found that the first Mitchell factor disfavors immunity for school districts. See, e.g., Holz, 347 F.3d at 1184 (Alaska); Savage, 343 F.3d at 1044 (Arizona); Eason, 303 F.3d at 1143 (Nevada). In Alaska and Nevada, for example, the state guarantees minimum funding for school districts — called the "basic support guarantee" in Nevada and the "basic need" in Alaska — and school districts are free to raise additional revenue beyond that amount. Holz, 347 F.3d at 1183-84; Eason, 303 F.3d at 1142-43. Because per-pupil spending need not be equalized across districts, we held it was "not necessarily true that an amount withdrawn from a school district's account in order to pay a judgment will be replaced with state money." Holz, 347 F.3d at 1184 (quoting Eason, 303 F.3d at 1143). Similarly, in Savage, we held the state of Arizona would not be liable for judgments against school districts, as districts' funds "are not subject to state control, are not subject to a Belanger-style spending-cap, and will not be replenished with money out of the state treasury." 343 F.3d at 1044.

Sato argues that when AB 97 replaced revenue limits with the LCFF, it replaced a maximum per-pupil funding formula with a minimum per-pupil formula. According to Sato, school districts and COEs in California, like those in Alaska, Arizona, and Nevada, are now free to raise local property tax revenues as much as they want to supplement minimum state support.

We find Sato's arguments without merit. As we noted in Belanger, equalization of per-pupil spending in California is not simply a matter of policy, but of state constitutional law. 963 F.2d at 251. In two decisions from the 1970s — Serrano v. Priest (Serrano I), 5 Cal.3d 584, 96 Cal.Rptr. 601, 487 P.2d 1241 (1971), and Serrano v. Priest (Serrano II), 18 Cal.3d 728, 135 Cal.Rptr. 345, 557 P.2d 929 (1976) — the California Supreme Court declared that disparities in per-pupil spending, driven by differences in districts' ability to generate local property tax revenues, violated the equal protection guarantees of the state constitution. Serrano II, 135 Cal.Rptr. 345, 557 P.2d at 957-58; Serrano I, 96 Cal.Rptr. 601, 487 P.2d at 1263. Nowhere in AB 97 does the California legislature attempt to amend the state constitution's equal protection provisions, and thus we assume that the constitutional precepts announced in Serrano I and Serrano II remain good law. See Methodist Hosp. of Sacramento v. Saylor, 5 Cal.3d 685, 97 Cal.Rptr. 1, 488 P.2d 161, 164-65 (1971) ("Unlike the federal Constitution, which is a grant of power to Congress, the California Constitution is a limitation or restriction on the powers of the Legislature.").

Even putting state constitutional issues aside, Sato's characterization of the LCFF as a minimum per-pupil funding formula is at odds with the text of AB 97, which demonstrates the legislature's intent to maintain equalization of per-pupil spending. Under the LCFF, districts and COEs receive a "base grant" amount for each student, and they may receive additional "supplemental grant" and "concentration grant" funding depending on the number of low-income and ELL students.6 Cal. Educ. Code § 2574(b)-(c) (COEs); id. § 42238.02(b)-(f) (districts). Supplemental grants provide school districts with additional funding equal to 20% of the base grant for each low-income or ELL student in the district; concentration grants provide an additional 50% of the base grant amount for each low-income or ELL student, if the proportion of low-income/ELL students exceeds 55% of total enrollment. Id. § 42238.02(e)-(f).

As explained by the California Legislative Analyst's Office, the combination of base, supplemental, and concentration grants ensures that "[d]istricts serving the same number of students in the same grade spans with the same characteristics receive the same amount of funding."7 Moreover, the state calculates its funding share to districts and COEs in the same manner as under the revenue limits system: the state's contribution is equal to the state-determined LCFF funding target minus local property tax revenue. Cal. Educ. Code §§ 2575, 42238.02(j). The legislature's use of the same state contribution equation indicates that the LCFF funding target, like a revenue limit, represents a maximum expenditure, not a minimum.

Sato contends that the legislature's use of the term "base grant" evinces its intent to allow districts and COEs to raise local property tax revenues above and beyond the state's minimum "base" support. He also alleges that OCDE receives a significant portion of its funding from local property tax revenues, which he claims are raised on top of LCFF base grant funding. But Sato's focus on semantics ignores the strictures of the California Constitution expressed in Serrano I and Serrano II. And there is nothing unusual about the fact that OCDE receives some of its funding from local tax dollars. As under the revenue limits system, school districts and COEs are funded through a combination of local tax revenues and the state's funding contribution. See Cal. Educ. Code. § 42238.02; Belanger, 963 F.2d at 252.

Examining the purposes of AB 97, we find no intent on the part of the California legislature to replace a maximum per-pupil funding formula with a minimum per-pupil formula. Rather, the legislation's objectives were twofold: (1) increase education funding,8 particularly for low-income and ELL students, and (2) simplify what many considered an "overly complex, inefficient, and outdated" education finance system.9 To achieve these ends, the LCFF streamlined the numerous, piecemeal sources of state funding under the old revenue limits system. Base grants, supplemental grants, and concentration grants replaced revenue limits funding and almost all of the dozens of categorical funding grants under the old funding mechanism. Taylor, supra note 8, at 6, 7 fig. 5 (listing 32 categorical funding programs eliminated under the LCFF). Categorical grants were restricted funds used for dozens of school services, such as special education, educational technology, and principal training. Now, instead of receiving many types of restricted funds, districts and COEs receive larger lump sums, with more flexibility over how state dollars are spent. See Campaign for Quality Educ. v. State, 209 Cal.Rptr.3d 888, 923 (Cal. Ct. App. 2016) (Liu, J., dissenting from denial of petition for review) (describing the LCFF as "designed to improve the ability of school districts to flexibly use state education dollars to address the most pressing areas of need").

While the passage of AB 97 provided districts and COEs with additional flexibility in their budgets, we note, however, that AB 97 did not eliminate the "centralized" system of "strict state control" over local districts' funding we discussed in Belanger. See 963 F.2d at 252. In Belanger, we noted that the enactment of Proposition 13 by statewide ballot initiative in 1978 "ensured that the state, rather than local school districts, would control funding for public schools." Id. at 251. Proposition 13 capped local property tax rates at 1% of assessed value, and transferred the authority to collect and redistribute local property tax revenues to the state. Cal. Const. art. XIIIA, § 1(a). As the California Supreme Court explained:

Cal. Redev. Ass'n v. Matosantos, 267 P.3d 580, 589 (Cal. 2011) (citations omitted). AB 97 did nothing to override Proposition 13, and Sato does not dispute that Proposition 13 remains in effect.

In sum, we find that AB 97 left in place the fundamental elements of Belanger: equalization of per-pupil spending and centralized control over local education budgets. Because the LCFF keeps in place a maximum per-pupil funding formula, state and local funds are still "hopelessly intertwined," and "any change in the allocation of property tax revenue has a direct effect on the allocation of state funds." Belanger, 963 F.2d at 252. The first Mitchell factor therefore weighs in favor of Eleventh Amendment immunity.

2

The second Mitchell factor asks us to consider "whether the [government] entity performs central government functions." Mitchell, 861 F.2d at 201. In Belanger, we found that "California law treats public schooling as a statewide or central governmental function." 963 F.2d at 253. The California Supreme Court has said that "[s]ince its admission to the Union, California has assumed specific responsibility for a statewide public education system open on equal terms to all." Butt v. State, 4 Cal.4th 668, 15 Cal.Rptr.2d 480, 842 P.2d 1240, 1247 (1992). Examples of the state's assumption of public education as a state responsibility appear throughout the California Constitution. See, e.g., Cal. Const. art. IX, § 5 ("The Legislature shall provide for a system of common schools."); Cal. Const. art. XVI, § 8(a) ("From all state revenues there shall first be set apart the moneys to be applied by the State for support of the public school system and public institutions of higher education.").

Sato argues that AB 97's requirement that school districts and COEs adopt Local Control and Accountability Plans (LCAPs) "unequivocally establish[es] that California public schooling is now controlled and provided locally." We disagree.

LCAPs are three-year plans that must describe districts' goals for student achievement and identify actions districts will take to meet those goals. Cal. Educ. Code § 52060; Taylor, supra note 8, at 8. LCAPs must address eight areas of state priority, and they may also include "additional local priorities identified by the governing board of the school district" or county board of education. Cal. Educ. Code §§ 52060(c)-(d), 52066(c)-(d). Within these areas, districts and COEs must set goals for all students and for specific subgroups of students, including ethnic minorities, low-income students, ELL students, and students with disabilities. Id. §§ 52052(a), 52060(c), 52066(c). LCAPs must be developed using a state-provided template. Id. §§ 52060(a), 52066(a). Within five days of adopting or updating its LCAP, a district must submit its LCAP to its COE for review, id. § 52070(a), and the COE is then responsible for filing the LCAP with the state, id. § 52070.5.

Sato argues that, in empowering districts to identify local priorities and develop strategies for achieving their goals through the LCAP process, AB 97 effectively transferred primary authority over public education from the state to the local level. We are unconvinced that LCAPs represent such a sea change. Under the state constitution, the provision of public education in California remains a central government function. And even through the LCAP process, the state still exercises significant control over school districts and COEs. Under the LCAP provisions of AB 97, districts and COEs "shall" address eight areas of priority identified by the state, and "may" also address local priorities. Cal. Educ. Code §§ 52060(c), (h), 52066(c), (h). The far-reaching areas of state priority include such fundamental issues as "[i]mplementation of the academic content and performance standards adopted by the state board," "[p]arental involvement," and "[p]upil achievement." Id. § 52060(d). For each area of state priority, districts and COEs must set goals for state-identified subgroups of students, and they must use state-provided templates and submit the final LCAP to the California Department of Education.

On the whole, AB 97 granted districts and COEs some measure of autonomy and discretion in goal-setting, but it did not delegate primary responsibility for providing public education. We therefore find that the passage of AB 97 did not disturb our longstanding precedent that "California law treats public schooling as a statewide or central governmental function." See Belanger, 963 F.2d at 253. As in Belanger, "[t]hat the state itself has decided to give its local agents more autonomy does not change the fact that the school districts remain state agents under state control." Id. The second Mitchell factor weighs in favor of Eleventh Amendment immunity for OCDE.

3

The passage of AB 97 does not impact our analysis of the remaining three Mitchell factors. As in Belanger, the third Mitchell factor — whether the government entity may sue or be sued — weighs in Sato's favor. See Belanger, 963 F.2d at 254. Under California law, COEs and school districts can sue and be sued in their own name. Cal. Educ. Code § 35162 ("In the name by which the district is designated the governing board may sue and be sued, and hold and convey property for the use and benefit of the school district."). Accordingly, this factor "militates against a finding of Eleventh Amendment immunity." Belanger, 963 F.2d at 254. But as we recognized in Belanger, "it does not necessarily follow that the school district can be sued for money damages just because it can be sued in its own name." Id. We therefore adopt Belanger's approach in finding that this factor weighs in Sato's favor, but we afford it less weight than the first two factors. See id.

4

The fourth Mitchell factor also weighs in Sato's favor. The fourth factor considers "whether the entity has the power to take property in its own name or only the name of the state." Mitchell, 861 F.2d at 201. In California, the same statute that allows COEs and school districts to sue and be sued in their own name also conveys to COEs and school districts the power to hold property in their own name. Cal. Educ. Code § 35162. Critically, however, the California Supreme Court has determined that local public school entities hold property for the benefit of the state. Hall v. City of Taft, 47 Cal.2d 177, 302 P.2d 574, 577 (1956) ("The beneficial ownership of property of the public schools is in the state."). We thus concluded in Belanger that this factor was a "close question" entitled to "little weight." 963 F.2d at 254. We reiterate that conclusion here.

5

The fifth Mitchell factor asks us to consider the corporate status of OCDE. In California, "school districts have the corporate status of agents of the state for purposes of school administration." Eason, 303 F.3d at 1144 (quoting Belanger, 963 F.2d at 254). The California Supreme Court has stated that "[l]ocal districts are the State's agents for local operation of the common school system." Butt, 15 Cal.Rptr.2d 480, 842 P.2d at 1248.

Sato contends that COEs must be a distinct corporate entity from the state, because OCDE sued the state of California in Orange County Department of Education v. California Department of Education, 668 F.3d 1052 (9th Cir. 2011). As we have already determined, however, California law previously authorized school districts and COEs to sue in their own name before the passage of AB 97. Cal. Educ. Code § 35162. This limited authority does not change the status of districts and COEs as agents of the state. This factor weighs in favor of Eleventh Amendment immunity.

III

Considering all of the Mitchell factors together, we hold that California school districts and COEs, including defendant OCDE, remain arms of the state and continue to enjoy Eleventh Amendment immunity. AB 97 reformed the financing and governance of California public schools in important ways, but it did not so fundamentally alter the relationship between COEs and the state as to abrogate our decision in Belanger. Sato's claims against OCDE were properly dismissed.

AFFIRMED. Each party shall bear its own costs.