This case involves four consolidated cases appealed from two decisions of the trial court on remedial questions following a judgment for the plaintiffs on liability. (Plaintiff Hailey’s case was consolidated with the three original actions and certified as a class action on 28 July 1997). On 21 July 1995, the trial court entered judgment for the plaintiffs and concluded they were entitled to receive additional dis*139ability benefits. This judgment was affirmed by our Supreme Court in Faulkenbury v. Teachers’ and State Employees’ Ret. Sys. of North Carolina, 345 N.C. 683, 483 S.E.2d 422 (1997). These cases, certified as class actions, challenged the way disability benefits were calculated under the Teachers’ and State Employees’ Retirement System of North Carolina and the North Carolina Local Governmental Employees’ Retirement System. The Supreme Court noted that members of the class were:
[A]ll persons who were receiving disability benefits in a lesser amount than they would have received had the law not been changed; persons who retired on service retirement who could have retired on disability retirement at higher rates if the law had not been changed; all living heirs, beneficiaries, or personal representatives of any estate of one who was receiving less as a disability retiree than he would have received if the law had not been changed; and who had not selected a designated survivor beneficiary; and all living heirs, beneficiaries, or personal representatives of the estate of a deceased survivor beneficiary who was receiving benefits pursuant to the election of an option by a deceased disability retiree.
Id. at 698, 483 S.E.2d at 431.
Thus, the plaintiffs include all class members who had been employed for more than five years as of 1 July 1982 and whose retirement and disability benefits were vested under either the Teachers’ and State Employees’ Retirement System of North Carolina or the North Carolina Local Governmental Employees’ Retirement System.
On 1 July 1982, the method of calculating disability benefit payments was changed so that the plaintiffs received less in disability payments than they would have received had they retired for disability prior to that date. All of the plaintiffs became disabled after 1 July 1982 and received benefits which were reduced from what they would have received if there had been no change in the law on 1 July 1982. Plaintiff Woodard died after the commencement of this action and his widow was substituted as a plaintiff.
In its decision, the Supreme Court held that the change in calculation of the plaintiffs’ disability benefits by the retirement systems impaired their contractual rights. Id. at 694, 483 S.E.2d at 429. According to the decision in Faulkenbury, the additional disability *140payments owed to the plaintiffs were to be determined by applying N.C. Gen. Stat. § 128-32 and N.C. Gen. Stat. § 135-10 which have virtually identical provisions as follows:
Should any change or error in the records result in any member or beneficiary receiving from the Retirement System more or less than he would have been entitled to receive had the records been correct, the Board of Trustees shall correct such error, and as far as practicable, shall adjust the payment in such a manner that the actuarial equivalent of the benefit to which such member or beneficiary was correctly entitled shall be paid.
The trial court held in its 21 July 1995 judgment that it “retainfed] jurisdiction ... to decide at a second trial, if necessary, the issues of specific amounts of underpayments, interest and actuarial interest due to each class member. ...”
In addressing this matter, the Supreme Court stated that the relevant statute sections showed it was the intent of the General Assembly that “if there was an underpayment of a pension compensation, it would be paid at the actuarial value.” Faulkenbury, 345 N.C. at 695, 483 S.E.2d at 430. Regarding the payment of interest on underpayments, the Supreme Court stated that if the state or local governments possessed sovereign immunity, it was waived by N.C. Gen. Stat. § 135-1(2), which defines actuarial equivalent as a “benefit of equal value when computed upon the basis of such mortality tables as shall be adopted by the Board of Trustees, and regular interest.” Id. at 696, 482 S.E.2d at 430 (quoting N.C. Gen. Stat. § 135-1(2) (1995)). Further, the Court noted that N.C. Gen. Stat. § 128-21(2) defines actuarial equivalent as a “benefit of equal value when computed at regular interest upon the basis of such mortality tables as shall be adopted by the Board of Trustees.” Id. (quoting N.C. Gen. Stat. § 128-21(2) (1995)). Also, the Supreme Court held that these sections require regular interest at four percent (4%) to be included in the actuarial value so the plaintiffs are entitled to the actuarial value of underpayments including interest. Id.
Thus, pursuant to Faulkenbury, the plaintiffs who were presently receiving disability benefits were entitled to pursue claims for underpayments for the three years before these actions were commenced. Faulkenbury, 345 N.C. at 695-96, 483 S.E.2d at 429-30. Further, in determining that plaintiffs were entitled to regular interest on the *141underpayments, the Court stated the following in addressing the defendants’ argument:
The defendants say that allowing recompense under all these sections gives the plaintiffs double recovery. They say that the payment of underpayments at their actuarial equivalent will fully compensate the plaintiffs and that the plaintiffs should not be paid interest. We disagree.
In allowing interest, the court was following the definition of actuarial equivalent prescribed by N.C.G.S. § 128-21(2) and N.C.G.S. § 135-1(2). There is no double recovery.
In its decision, the Supreme Court remanded the case to the trial court for a determination of the additional disability benefits due the plaintiffs.
At the hearing in the trial court, the evidence consisted of the deposition testimony of the parties’ respective actuarial experts. The plaintiffs offered the deposition testimony of their expert, Robert G. Sanford Jr. (Sanford), along with various exhibits in support of his calculations. The defendants offered the deposition testimony of their expert, Donald M. Overholser (Overholser). In arriving at his conclusion of the additional disability benefits due the plaintiffs, Sanford utilized a calculation of actuarial equivalent which required the inclusion of a mortality factor. The defendants’ expert, Overholser, calculated the additional disability benefits without the application of a mortality factor.
On 25 August 1997, the trial court found that Sanford’s “calculations and methodology are in accord with the statutory definition of actuarial equivalent and contain the correct calculation and methodology for calculating pension underpayments” due the plaintiffs. The trial court then ordered that the additional disability benefits due the plaintiffs be “calculated by application of the statutory definition of actuarial equivalent” based on a mortality factor as testified to by Sanford. On 24 October 1997, another hearing was held after which the trial court entered an order upholding the plaintiffs’ claim for judgment interest and denied the defendants’ motion for a stay of the order of 25 August 1997.
On appeal, where the facts are not at issue, a de novo standard of review is applied in determining whether an error of law exists. Ayers v. Bd. of Adjust. for Town of Robersonville, 113 N.C. App. 528, 530, *142439 S.E.2d 199, 201, disc. review denied, 336 N.C. 71, 445 S.E.2d 28 (1994); Capricorn Equity Corp. v. Town of Chapel Hill, 334 N.C. 132, 431 S.E.2d 183 (1993). “Any error made in interpreting a statute is an error of law.” Savings & Loan League v. Credit Union Comm., 302 N.C. 458, 464, 276 S.E.2d 404, 409 (1981).
 On appeal, the defendants contend that the trial court erred as a matter of law in requiring the payment of additional disability benefits based on a mortality factor because it mistakenly interpreted the statutes governing the retirement system and the Supreme Court’s decision in Faulkenbury. The defendants further contend that the use of a mortality factor is not relevant in the calculation of additional disability benefits owed to the plaintiffs or their survivors as there is no risk of such benefits being forfeited. On the other hand, the plaintiffs contend that both N.C. Gen. Stat. § 128-32 and § 135-10 require that the actuarial equivalent based on a mortality factor be used to calculate the amount of additional disability benefits to be paid as determined by the decision in Faulkenbury. Furthermore, the plaintiffs argue that since they faced the risk of dying before having received the additional disability benefits, the use of a mortality factor is relevant in making the calculations.
The plaintiffs’ expert, Sanford, testified that he is employed as a consulting actuary with ADP Benefit Services in Richmond, Virginia, where he has 20 years of experience in the design, financing, management, and administration of employee benefit programs. Sanford noted that he had been contacted by the plaintiffs’ counsel and informed “that there had been a decision that a group of retirees were entitled to be caught up” and paid “past underpaid retirement benefits.” He stated that he was “hired to perform the calculation of the amount needed to make up for those past benefits.” In his approach to this case, Sanford stated that he reviewed the statutes applicable to the retirement systems in Chapters 135 and 128 and the Supreme Court’s decision in Faulkenbury, all of which were furnished to him by the plaintiffs’ counsel. He explained why his approach included a mortality factor in his calculations in addition to the regular interest.
On direct examination, he testified:
Q. In your- opinion what is missing from a calculation made . . . using only the four percent interest component?
*143A. This plan’s definition defines two components in computing the actuarial equivalent. One is the regular interest and the other is the prescribed mortality tables. ...
Sanford then gave an explanation of “actuarial equivalent” as follows:
The Actuarial Equivalent of a benefit is the amount of money that a benefit is ‘worth’ at a given date based on a stated mortality table and interest rate. If the Actuarial Equivalent of a benefit is determined at the beginning of the benefit payout period, it is determined by discounting the future benefit payments for both interest and mortality, i.e. the probability that the person will survive to receive each future benefit [payment]. If the Actuarial Equivalent of a benefit is determined at the end of a payment period, the reverse is true. The ‘accumulated’ Actuarial Equivalent must accumulate prior benefits with interest over the past time period and also for the inverse of mortality, or the ‘benefit of survivorship.’
Further, on direct examination, Sanford was asked to illustrate the actuarial equivalent with a hypothetical where a person’s benefit is incorrect at age 50 and corrected at age 55:
Q. If you will, illustrate for us . . . what it is you are talking about when you say actuarial equivalent with reference, say to a person 50 years old and, say, for over a period of five years. Illustrate . . . with reference to that person where the benefit is incorrect at age 50 and is corrected — when that person reaches the age of 55.
A. We are dealing with a single benefit of $1,000 that was payable at age 50. And we are going to accumulate that for mortality and 4 percent interest. The interest accumulation would — just using simple interest on the original amount, would accumulate $40 every year.. . .
The reflection of the mortality component requires that you select — you construct what we call a life table based on these mortality rates. You start with one person in your life table at age 50 . . . let’s assume that every year the chance of dying is 1 percent. At age 51, we are down to .99 of a person, .98,.97,. . . .
*144The actuarial equivalent of the $1,000 is your original $1,000 plus your $200 interest divided by the .95. There is a 5 percent chance of dying over the five year period.
Q. And so the bottom line answer to the equation you have on the bottom there, then, would give you the actuarial equivalent of what that person should have gotten over that five year period.
On cross-examination, Sanford was questioned further about how he defined the “benefit of survivorship.”
Q. And what do you mean by the benefit of survivorship?
A. The benefit of survivorship is this mortality factor we went through here where you — it is basically the probability of having died from the time those payments were made until the present time.
Q. What is the benefit of survivorship in this instance?
A. . . . [I]t is a calculation that, I believe is called for under the plan, that in addition to the interest component of actuarial equivalent it prescribes that mortality be reflected in the calculation.
Q. Well, why would you reflect the benefit of survivorship if whether the person lives doesn’t make any difference?
A. Because the plan definition calls for it.
In referring to his hypothetical of an incorrect benefit being paid to a person at age 50, but corrected at age 55, Sanford testified as follows:
Q. And why is that — can you tell us simply why that is the value of it?
A. The interest component is just a simple accumulation on the $1,000 for five years. The mortality component — it shows here that the probability of dying in that five year period is 5 percent. And that is the survivorship piece that gets added on.
Q. And the significance there is that some people die and do not receive their money?
*145A. Right. The concept is, you know, a little bit like you are accumulating a retirement fund and everybody is setting aside money, but some people will not live to ever receive anything. And part of the funding for the people who do survive will come from the money that is forfeited by the people that die in the meantime. That is, you know, somewhat of an idea of this benefit of survivorship.
In summary, Sanford testified that his approach of using a mortality factor was based on the usual assumption of a risk that benefits would not be paid to a survivor of a member and thus would be forfeited. While the plaintiffs argue that they faced the risk of dying before having received the additional disability benefits, Sanford stated that he included the benefit of survivorship in his calculations because he understood the plan called for it. He admitted his calculations were based on benefits not being paid to a survivor of a member.
The defendants’ expert, Overholser, testified that he is employed at Buck Consultants where he is the principal consulting actuary and works primarily with governmental retirement systems. He serves as director of the public employees’ retirement system consulting practice for the firm and is a member of both task forces that worked with the governmental accounting standards board in setting their standards for public pension plans. He is also the chairman of the subcommittee on public retirement systems of the Academy of Actuaries. In addition, he currently serves as the actuary for the statewide retirement systems in South Carolina, Georgia, and Alabama, as well as for the teachers’ retirement system in Kentucky, and has served as actuary for the retirement systems in North Carolina since 1985.
Overholser explained that he examined the same information as Sanford, including the Faulkenbury decision, relevant statutes, and mortality tables. On direct examination he was asked:
Q. [W]hat is an actuarial equivalent or what is it supposed to be?
A. It is a payment of equivalent value. Mr. Sanford in his testimony gave a definition of actuarial equivalent, which I would say is pretty accurate. It is just converting one stream of payments into another stream of payments so that they are equivalent when taking into account the time value of money and the chance or probability that a person will receive payment or a series of payments.
*146In all actuarial calculations that deal with equivalency, three factors are considered: the benefits to be paid, the time value of money factor, and the survivorship factor are always combined and you get a value on that basis. Doing this you can always equate streams of payments. This is kind of what the concept of actuarial equivalent is.
In his capacity as a consulting actuary for this State’s retirement systems, he stated that actuarial equivalent factors were regularly used.
Q. And when would they use them?
A. When members retire, if they elect an optional form of benefit which is the actuarial equivalent of their basic benefit, the factors are used to convert the regular allowance into the optional allowance, into the actuarial equivalent of the optional allowance.
Q. [W]ould that be, for example, if someone retired at age 65 and then instead of taking the maximum benefit he chose to take a smaller benefit so that his wife could have a benefit after his death? Is that when you use this calculation?
A. The kind you just described is a joint survivor benefit or a sur-vivorship benefit where you leave a benefit to a survivor after your death.
Overholser then testified that he disagreed with the calculations by Sanford. He explained his disagreement in the following manner:
I disagree with the calculations in the context of the . . . decision of the court in this case.
These factors and this calculation would be appropriate if we were only going to make payments to the members who have survived until the present time. These calculations build in a sur-vivorship factor when there is really no contingency risk or forfeiture of the benefits.
According to the court ruling, not only the people who survive would be entitled to payment, but also the people who have died since they received the payments and before the present time. So there is really no risk of forfeiture; hence there is no sur-*147vivorship value involved. Hence the mortality or survivor factor would not be involved in these calculations.
Overholser was further asked to explain how he interpreted the language of the retirement statutes to his determination that a mortality factor is not relevant in the calculation of additional disability benefits here.
Q. And how would you relate the language in the statutory provision about the mortality tables to your opinion that no calculation for mortality should be included?
A. Well, in calculating actuarial equivalency, you apply a mortality table if it is relevant to the calculation, in particular if there is a benefit of survivorship involved. If there is no benefit of sur-vivorship, you do not introduce the mortality element. . .. Where there is an element of risk of forfeiture involved, . . . you do reflect the mortality. . . .
There is nothing at all contradictory in not using mortality here because it doesn’t enter into an actuarial equivalent calculation when there is no element of forfeiture or risk.
Q. Are you aware of any experience or any instance in which mortality tables are used to calculate an actuarial equivalent when the risk of mortality does not prevent payment.
A. I am not.
In summary, Overholser explained that to include a mortality factor in the calculation when there was no risk of losing benefits would be to create “value out of nothing.” He stated Sanford’s calculations would be correct if you have to survive to get the benefit, and those who do survive would get a “windfall, a benefit of survivorship, because [they] profit from the forfeiture of benefits by the people who have died in the interim.” However, Overholser concluded that since the Court had ordered that the additional disability benefits be paid “to people who have died ... there is no benefit of survivorship] from forfeiture.”
Therefore, it is apparent from the testimony of Sanford and Overholser that the foundation of the mortality factor used in the calculation of a retroactive benefit is premised on whether that benefit is forfeited by a member who does not survive to the time of payment and whether that payment which is forfeited is paid to those members who have survived from the original group.
*148Black’s Law Dictionary defines mortality tables as, “A means of ascertaining the probable number of years any man or woman of a given age and of ordinary health will live.” Black’s Law Dictionary, 1009 (6th ed. 1991). “A mortality table should only be used for the purpose it is meant to fulfill.” Vinson & Elkins v. Commissioner, 99 T.C. 9, 53 (1992), affirmed, 7 F. 3d 1235 (1993) (explaining the determination of which mortality table was to be used in determining pre-retirement death benefits versus post-retirement annuities). The key factor in determining when to apply a mortality factor is whether or not survivorship is a risk factor in the calculation of benefits to be paid. As Sanford explained in his testimony, the benefit of surviv-orship is “the probability of having died from the time those payments were made until the present time” with those payments being forfeited.
Within the retirement statutes a mortality factor is relevant, as Overholser pointed out, when, for example, an employee retires, instead of taking the maximum benefit, he chooses to take a smaller benefit so that his beneficiary can have a larger benefit after the employee’s death. In this example, the life expectancy of the employee determines the amount of the benefit because it affects the likelihood that the benefit will continue to a certain point. The use of mortality tables may be ignored if a forfeiture of compensation does not occur at death. William H. Schmidt, Limitations on Contributions and Benefits, C529 A.L.I.-A.B.A. 137 (1990) (explaining the use of mortality tables in determining the actuarial equivalence of a defined benefit plan). “[I]t is a well settled rule of statutory construction that, where a literal interpretation of the language of a statute would contravene the manifest purpose of the statute, the reason and purpose of the law will be given effect and the strict letter thereof disregarded.” Matter of Banks, 295 N.C. 236, 240, 244 S.E.2d 386, 389 (1978).
The plaintiffs argue they faced the risk of dying while awaiting the underpayments while Sanford’s calculations were based on the risk of forfeiting these payments by members who had died. Our retirement statutes do not recognize the risk asserted by the plaintiffs and included in Sanford’s calculations. Since there is no forfeiture of payments by deceased members, there are no other risks associated with these underpayments.
We have carefully reviewed all relevant statutes and the decision in Faulkenbury including the option elections for retirees and we *149find no authority which mandates the calculation of additional disability benefits as contended by the plaintiffs.
The trial court construed the Supreme Court’s decision in Faulkenbury to mandate the use of a mortality factor in computing the actuarial equivalent of the additional disability benefits. We do not construe our retirement statutes and the decision in Faulkenbury to mandate such a construction. The trial court erred as a matter of law in ordering that the additional disability benefits due plaintiffs be calculated according to the plaintiffs’ expert.
 Next, the defendants contend that the trial court erred in allowing the plaintiffs to collect post-judgment interest on their retroactive benefits. The trial court awarded post-judgment interest at the legal rate of eight percent (8%), as set by N.C. Gen. Stat. § 24-1 (1991), in addition to the regular interest of four percent (4%) required under the retirement statutes.
The plaintiffs contend that they should be granted post-judgment interest pursuant to our Supreme Court’s decision in Smith v. State, 289 N.C. 303, 222 S.E.2d 412 (1976). In that decision, the Supreme Court held that when the State enters into a contract, it implicitly agrees to be sued for damages. The Court further held that when the State is sued for damages, it will “occupy the same position as any other litigant.” Id. at 320, 222 S.E.2d at 424. Therefore, the plaintiffs argue that if the State is to be treated like “any other litigant,” that means it should be required to pay post-judgment interest. However, the defendants contend that despite the holding in Smith, the State is not required to pay post-judgment interest unless there is statutory authority requiring it to do so.
Our Courts have held that the State is not required to pay interest on its obligations unless it is required to do so by contract or by statute. Stanley v. Retirement and Health Benefits Division, 66 N.C. App. 122, 123, 310 S.E.2d 637, 638, disc. review denied, 310 N.C. 626, 315 S.E.2d 692 (1984); Davidson & Jones, Inc. v. N.C. Dept. of Administration, 69 N.C. App. 563, 570, 317 S.E.2d 718, 723, affirmed in part and reversed in part, 315 N.C. 144, 337 S.E.2d 463 (1985); Yancey v. Highway Commission, 222 N.C. 106, 109, 22 S.E.2d 256, 259 (1942). This has been the long established rule in this State. The decision in Smith did not change this rule. In Smith, the issue of whether to allow post-judgment interest was not before the Court. Instead, that case dealt with the issue of whether a plaintiff could sue the State in a breach of contract action. Smith, 289 N.C. at 320, 222 *150S.E.2d at 424. The decisions by our appellate courts since Smith confirm this point. See Stanley, 66 N.C. App. at 123, 310 S.E.2d at 638; See Davidson & Jones, Inc., 69 N.C. App. at 570, 317 S.E.2d at 723.
Under our retirement statutes, N.C. Gen. Stat. § 135-1(2) and § 128-21(2), regular interest is to be included in the payment of retroactive retirement benefits. Statutes that allow recovery against the State are to be strictly construed. Myers v. Dept. of Crime Control, 67 N.C. App. 553, 555, 313 S.E.2d 276, 277 (1984). In Myers, this Court held that the plaintiff was not entitled to post-judgment interest on an award of damages under the State Tort Claims Act. Id. This Court, following the reasoning in Yancey, stated that post-judgment interest is not collectible against the State without authorization by the General Assembly or unless the State has agreed to do so. Id. The Court further noted that interest can be assessed at the legal rate on recovery of workers’ compensation benefits because N.C. Gen. Stat. § 97-86.2 provided that such interest was to be awarded. Id. at 555-56, 313 S.E.2d at 277. However, N.C. Gen. Stat. § 24-1 et seq., which allows for post-judgment interest, contains no provision for the allowance of such interest to be awarded against the State. The General Assembly has not authorized the allowance of post-judgment interest against the State but has provided that all retirement benefits shall include regular interest of four percent (4%). Therefore, post-judgment interest should not have been awarded against the defendants.
In conclusion, the order of the trial court is reversed and the case is remanded for a determination of additional disability benefits owed to the plaintiffs consistent with this opinion.
Reversed and remanded.
Judges LEWIS and MARTIN, John C., concur.