retired superior court justice, specially assigned under RSA 490:3, concurring in part and dissenting in part. I agree with the majority that the prior retirement statutes created a contract between the State and the petitioners and that RSA chapter 100-C impairs that contract. I also agree that the salary adjustments provided for by Laws 2003, 311:3 and Laws 2005,177:96 should not be included in calculating benefits under the prior retirement statutes. However, I disagree that the case should be remanded for further findings by the trial court as to whether the impairment of the petitioners’ contract rights is offset by any compensating benefits under the new law. For the reasons that follow, I would affirm the trial court’s finding of substantial impairment. I would also affirm the trial court’s finding that the new statute is not reasonable and necessary to serve an important public purpose, an issue which the majority does not reach.
The prior retirement statutes tied the petitioners’ retirement benefits directly to salaries of sitting judges; namely, “3/4 of the currently effective annual salary of the office from which [the judge] is retired...” RSA 490:2, II (1997) (repealed 2003); RSA 491:2, II (1997) (repealed 2003); RSA 502-A:6-a, III (1997) (repealed 2003). RSA chapter 100-C cuts that link and instead fixes the petitioners’ retirement benefits at “75 percent of the member’s final year’s salary,” RSA 100-C:5, II, subject to discretionary increases by the Board of Trustees of the New Hampshire Judicial Retirement Plan (board), RSA 100-C:13,111(g); RSA 100-C:17. As demonstrated by the variety of projected payout calculations in the parties’ joint stipulation, it is not possible precisely to quantify the difference between the financial benefits to be expected under the new and old retirement *459statutory schemes. However, it cannot be denied that the new system allows for the possibility that the benefits paid to retired judges will not keep pace with the salaries of sitting judges. That possibility did not exist under the prior laws. Thus I agree that RSA chapter 100-C impairs the State’s obligations entered into under the prior retirement statutes.
In seeking to distinguish this case from Opinion of the Justices (Furlough), 135 N.H. 625 (1992), the majority notes that the old system did not guarantee any specific increases to the retired judges (just as the new system does not), and that, in eight of the years between 1991 and 2010, there was no increase in judicial salaries. In my opinion, this misses the point. Under the old system, the petitioners could count on automatic increases whenever sitting judges received them. Under the new system, that is no longer true. No longer can they trust in the fact that the legislature will inevitably, albeit not at regular intervals, increase the salaries of sitting judges as dictated by the need to attract qualified persons to the bench. Instead, they must rely solely on the discretion of the board, which is subject to additional statutory preconditions described in the majority opinion. As the petitioners put it, “retired judges’ COLA’s are, under the prior system, a function of the employment market because the retired judges remain in 75% parity with sitting judges’ salaries. COLAs under RSA chapter 100-C, on the other hand, are simply a function of the Board’s discretion, driven largely, if not entirely, by the Plan’s recent and forecast investment returns.” The trial court acknowledged this difference between the two plans, finding that the new law “prohibits retired judges from receiving the advantages of any raises or COLAs instituted for the benefit of the judges presently sitting after their retirement.”
Thus, while I agree with the majority that the contract right impaired in this case does not have the degree of definiteness as the right at issue in Opinion of the Justices (Furlough), 135 N.H. at 634, I still think it clear that RSA chapter 100-C’s provision for discretionary increases is neither fair nor equivalent compensation for the impairment of the petitioners’ contract rights under the former plan. See Sylvestre v. State, 214 N.W.2d 658, 666 (Minn. 1973) (“Surely, if there is a contract, taking away the right to receive compensation based on the increased salary of judges without any compensating benefits constitutes an impairment of the contract.”); Wagoner v. Gainer, 279 S.E.2d 636, 644 (W. Va. 1981) (“other courts, in what we consider the better view, will allow a change in the plan if any disadvantages are counter-balanced by advantages, ie., substitute consideration”).
In Tuttle v. New Hampshire Medical Malpractice Joint Underwriting Association, 159 N.H. 627 (2010), this court recognized that “the determination of whether a contract impairment is substantial may be influenced *460by whether the contracting parties relied on the abridged contract right. Where the right abridged was one that induced the parties to contract in the first place, a court can assume the impairment to be substantial.” Tuttle, 159 N.H. at 649 (quotation omitted). In explaining its conclusion that the impairment to the contract was substantial, the trial court cited the foregoing passage. In the instant case, as in Tuttle, there is no dispute as to the meaning of the statutory provisions underlying the petitioners’ Contract Clause claim. Also as in Tuttle, the State did not contest the petitioners’ reliance upon the alleged contractual right at issue. Rather, the State argued at length that the statute did not create a contract and, in the alternative, that any contract that was created was not substantially impaired.
I agree with the trial court that the right to a retirement benefit tethered to the salary of currently sitting judges was an inducement upon which the petitioners reasonably relied in accepting the offer of appointment to the office. This conclusion is supported by the language of the statutes in effect when the petitioners accepted employment. Those statutes provided that the retirement benefit was “additional compensation for services rendered and to be rendered.” RSA 502-A:6-a, III (1997) (repealed 2003) (emphasis added); RSA 490:2, II (1997) (repealed 2003) (emphasis added); RSA 491:2, II (1997) (repealed 2003) (emphasis added). Retired judges continue to serve as judges and referees after they retire. RSA 490:3, II (2010); RSA 493-A:l, II (2010); RSA 493-A:l-a, I (2010); RSA 502-A:6-b (2010). That they have done so and continue to do so as a matter of practice is not subject to legitimate dispute. This distinguishes them from other public retirees and adds force to the argument that, in accepting appointment as judges, they were induced by and relied upon the statutory promise of retirement benefits specifically pegged to the salaries of sitting justices.
Under these circumstances, I would hold that RSA chapter 100-C, by cutting the link between the petitioners’ retirement benefits and the salaries of sitting justices and subjecting them to the discretion of the board without regard to the current salaries of sitting justices, substantially impairs the contract established by the retirement statutes in effect when the petitioners took office.
The majority opinion questions the trial court’s use of the inducement- and-reliance passage from Tuttle on two grounds, neither of which I find persuasive. As I read its opinion, the Tuttle majority did not conclude that reliance itself was irrelevant to its conclusion; rather, it deemed that reliance was established for purposes of the appeal. Tuttle, 159 N.H. at 649-50. Even though the Tuttle majority did not provide an analytical framework for deciding whether particular facts amount to reliance, its *461statement of the presumptive significance of reliance clearly contributed to its resolution of the issue of substantial impairment and thus is not mere dicta, as the majority maintains.
The majority also questions the soundness of Tuttle’s reliance-and-inducement rule, given that the language was taken from a district court opinion which was subsequently reversed by the Fourth Circuit. Fraternal Order of Police Lodge No. 89 v. Prince George’s County, Md., 645 F. Supp. 2d 492 (D. Md. 2009), rev’d, 608 F.3d 183 (4th Cir. 2010). However, the appellate court did not disagree with or even comment upon the district court’s substantial impairment analysis; it reversed because it found no contract impairment in the first place. Fraternal Order, 608 F.3d at 190-91. Moreover, the Fraternal Order district court opinion is not the sole authority supporting the importance of inducement and reliance in the context of a substantiality analysis. See, e.g., Houlton Citizens’ Coalition v. Town of Houlton, 175 F.3d 178, 190 (1st Cir. 1999) (“In order to weigh the substantiality of a contractual impairment, courts look long and hard at the reasonable expectations of the parties.”); Sal Tinnerello & Sons, Inc. v. Town of Stonington, 141 F.3d 46, 53 (2d Cir.), cert. denied, 525 U.S. 923 (1998) (“[T]he primary consideration in determining whether the impairment is substantial is the extent to which reasonable expectations under the contract have been disrupted.”). Both of these cases are cited in Tuttle. Tuttle, 159 N.H. at 669 (Dalianis and Duggan, JJ., dissenting). Although the dissent disagreed with the majority’s view that actual reliance was not at issue, it clearly shared the majority’s view that reliance was a primary consideration in determining whether a contract impairment was substantial.
Because I agree with the trial court’s finding of substantial impairment, I must also address the question whether the impairment of the petitioners’ contract rights was reasonable and necessary to serve an important public purpose.
“If the legislation substantially impairs the contract, a balancing of the police power and the rights protected by the contract clause[] must be performed, and the law may pass constitutional muster only if it is reasonable and necessary to serve an important public purpose.” Tuttle, 159 N.H. at 641 (quotations, ellipses, and brackets omitted).
We generally defer to the judgment of the legislature in determining whether a particular act is reasonable and necessary to serve an important public purpose, but when the State attempts to abrogate its own contractual responsibilities, complete deference to a legislative assessment of reasonableness and necessity is not appropriate because the State’s self-interest is at stake. Application of stricter judicial review reflects the principle that those who *462lawfully contract amongst themselves must have reasonable assurances that their rights and obligations will not be disturbed.
Lower Village Hydroelectric Assocs. v. City of Claremont, 147 N.H. 73, 78 (2001) (citations and quotations omitted). In this case, where the State was a party to the contract that is substantially impaired by RSA chapter 100-C, complete deference is not appropriate.
I agree with the trial court that the State’s justification for the contract impairment here was insufficient. The State argued that the changes at issue provided a long term fiscal solution to eradicate the unfunded liability created by the previous retirement statutes and that the changes also allowed an opportunity for benefits upon early retirement, which was not available previously. While the proffered justification may be reasonable going forward, there was no showing that it was reasonable and necessary for the legislature to apply these changes retroactively to the determinate class of judges who had accepted their appointments and served in reliance upon the provisions of the prior retirement system. See Miles v. Tenn. Consol. Retirement System, 548 S.W.2d 299, 305 (Tenn. 1976) (legislature did not have power to modify pension benefits “in the absence of a showing that a vital interest of the State must be protected by an exercise of the police power”).
Accordingly, I respectfully dissent from the majority’s decision to remand the case for further findings regarding substantiality of impairment. I would affirm the decision below.
Bean, J., retired superior court justice, specially assigned under RSA 490:3, joins in the opinion of Manias, J.