concurring.
Although I concur with the judgment, I cannot join the majority’s reasoning for three reasons: (1) its extension of Regents of the University of California v. Doe, 519 U.S. 425, 117 S.Ct. 900, 137 L.Ed.2d 55 (1997), (2) its focus on Ohio Rev.Code § 5101.16, and (3) its reliance on Ohio Rev. Code § 5101.161.
First, I have substantive concerns with applying and extending Doe to this case. Unlike Doe, which involved extinguishing a state’s immunity because of a voluntary third-party indemnification agreement, this case involves exposing the state to liability via a statutorily-mandated indemnification agreement. See generally Cash v. Granville County Bd. of Educ., 242 F.3d 219, 221 n. 1 (4th Cir.2001) (“And in Regents, the Court held that the fact that a judgment against the State would be covered by the voluntary indemnification agreement of a third party did not strip away the State’s Eleventh Anendment immunity because the State still bore the legal ‘risk of an [sic] adverse judgment.’ ”) (quoting Doe, 519 U.S. at 431, 117 S.Ct. 900). Rather than focus on the method of payment, the majority should have focused on whether a judgment “ ‘would have the practical effect of requiring payments from [Ohio].’ ” McGinty v. New York, 251 F.3d 84, 99 (2d Cir.2001) (quoting Mancuso v. N.Y. State Thruway Auth., 86 F.3d 289, 296 (2d Cir.1996)) (emphasis added). Using this approach, if “the vulnerability of the State’s purse is the most salient factor’ ” then it makes little difference to the state if it pays for a judgment directly or is statutorily required to reimburse the Hamilton County Department of Job & Family Services (“HCJFS”) for the same amount. Id. at 100 (quoting Hess v. Port Auth. Trans-Hudson Corp., 513 U.S. 30, 48, 115 S.Ct. 394, 130 L.Ed.2d 245 (1994)). Either way the state is required by law to distribute money from its treasury.
Extending Doe to the case at hand is unsupported by any case law, and at least one district court has rejected the majority’s interpretation of Doe. See Treasurer of State of Conn. v. Fortsmann Little & Co., No. 3:02CV519(JBA), 2002 WL 31455245, at *5 (D.Conn. Oct. 15, 2002). Even the Court’s opinion in Doe itself suggests that its holding was not intended to be em*333ployed this broadly: “The narrow question presented by this case is whether the fact that the Federal Government has agreed to indemnify a state instrumentality against the costs of litigation, including adverse judgments, divests the state agency of Eleventh Amendment immunity. We hold that it does not.” Doe, 519 U.S. at 426, 117 S.Ct. 900 (emphasis added); see also Kirchmann v. Lake Elsinore Unified Sch. Dist., 83 Cal.App.4th 1098, 100 Cal. Rptr.2d 289, 298 (2000) (emphasizing Doe’s reference to the question presented as “ ‘narrow’ ” and concluding that “Doe thus stands merely for the proposition that the fact state funds are not actually used to pay a judgment for which the state is legally liable does not preclude immunity”). Similarly, this application of Doe conflicts with our decision in Ernst v. Rising, 427 F.3d 351, 355 (6th Cir.2005) (en banc). In Ernst, we noted: “[The retirement system] is funded by the state treasury as well as by contributions from state officials. And if the retirement system faces a monetary shortfall, state legislation requires the state treasurer to make up the difference with state funds.” Id. (emphasis added). Likewise, in this case “state legislation requires” the state to reimburse HCJFS if the lawsuit is an administrative cost. Id. The majority neither explains how the reimbursement relationship in Ernst is any different than the one in this case, nor cites a single case to support its interpretation of Doe, despite the fact that it has been law for over a decade.
From a practical standpoint, I am concerned that expanding Doe will have unintended consequences regarding the way states structure their financial relationships with other state entities because we do not know how many state entities have similar payment relationships. This concern is compounded by the fact that the state is not a party in this suit and thus has not had a chance to weigh in on this issue. Notably, the majority opinion explicitly acknowledges that it does need to use Doe to reach its result: “even if we were to accept HCJFS’s reimbursement theory as determinative, HCJFS has not met its burden to show that the state necessarily will reimburse it for the damages awarded to Lowe.”
Second, the majority’s focus on Ohio Rev.Code § 5101.16 is misguided. This section establishes that the county must pay for a portion of HCJFS’s costs that are submitted for reimbursement, but the potential amount that must be paid appears to be no more than 10%, depending on certain financial criteria such as the income level of the county’s constituents. The majority cites no case law to support the conclusion that an entity must be “fully funded” to be an arm of the state, rather than receiving 90% of its funds from the state. On the contrary, persuasive authority indicates this factor is not relevant. See Hadley v. N. Ark. Cmty. Technical Coll, 76 F.3d 1437, 1440-41 (8th Cir.1996) (“The relevant funding inquiry cannot be whether NACTC enjoys some non-state funding, such as user fees (tuition), because then most state departments and agencies, and all state universities, would be denied Eleventh Amendment immunity.”). Furthermore, it is unclear why being “fully funded” is determinative since, depending on the judgment amounts, Ohio’s treasury can be more financially exposed even if it has to pay a lower percentage of the overall judgment.
Third, I am concerned about the majority’s reliance on Ohio Rev.Code § 5101.161, which may establish a threshold above which the state can deny reimbursement. The record reveals little about HCJFS’s reimbursement relationship with the state, and the majority has not established that this threshold exists in a material way, *334that it is enforced, that HCJFS would be in danger of surpassing this threshold if it loses this case, or that the state would not reimburse HCJFS if it surpassed the threshold because of a damages award in this case or for another reason. These factors indicate that HCJFS has failed to meet its burden of proof on this issue, but § 5101.161 itself should not preclude HCJFS from obtaining immunity without more information about how the state enforces the statute or some case law supporting the majority’s conclusion. See generally Hadley, 76 F.3d at 1441 (“The district’s never-exercised authority to supplement NACTC’s operating budget with limited local tax revenues does not change the fact that the State has created an institution of higher learning ‘that is dependent upon and functionally integrated with the state treasury.’ ”) (quoting Kashani v. Purdue Univ., 813 F.2d 843, 846 (7th Cir.1987) (emphasis added)).
Instead, the opinion should have focused on its conclusion that HCJFS failed to meet its burden of proof regarding whether the damages would be reimbursable as an administrative cost. HCJFS has not put forward sufficient evidence that the state would reimburse any litigation costs, let alone the costs that would be imposed if the plaintiff prevails in this case. By narrowly disposing of this case on the issue of what constitutes an administrative cost, this court could have avoided the unnecessary risk of sweeping too broadly with its interpretation of Doe, § 5101.161, and § 5101.16.