[Note: The following is a blog post I wrote for Reason.org, which you can also find here. Click here to see an archive of the posts I’ve written for Reason.org, including various articles about the law of public-employee pensions (this one and this one, as well as this white paper about the California Rule).]

* * *

Underfunded pension liabilities are a major fiscal problem for state and local governments. But as all public employee pension reformers are aware, the federal and state constitutions impose limits on how drastically public pension benefits can be cut. The chief relevant constitutional provision is the federal Constitution’s Contract Clause: “No State shall . . . pass any . . . Law impairing the Obligation of Contracts.” The Supreme Court judges state contractual impairments under a three-part test, stated in, among other cases, Energy Reserves Group, Inc. v. Kansas Power & Light Co. (1983): (1) Has there been a “substantial impairment of a contractual relationship”? (2) If so, does the state have a “significant and legitimate public purpose behind the regulation”? (3) If so, is “the adjustment of the rights and responsibilities of contracting parties . . . based upon reasonable conditions and . . . of a character appropriate to the public purpose justifying the legislation’s adoption”?

States generally have analogous language in their own constitutions: for instance, the Texas and Georgia Constitutions have Contract Clauses with nearly identical language to the federal version.

Despite this general uniformity, the constitutional law of public employee pension reform actually differs quite a bit from state to state. One reason is that some states have special, extra-protective provisions in their own constitutions, sometimes specifically aimed at pension protection. In Illinois, for instance, the state constitution provides that “[m]embership in any [state or local] pension or retirement system . . . shall be an enforceable contractual relationship, the benefits of which shall not be diminished or impaired”; the Arizona Constitution has nearly identical language, and the Michigan Constitution’s approach is similar.

Another reason is that states differ in whether they classify public employee pensions as contracts at all. Generally, public employee pensions are defined by statute rather than being written down in a physical contract. And classically, they were usually categorized as “gratuities”—which the state could change virtually at will, without running into any Contract Clause problems. Arkansas still takes this approach, at least for pensions with no employee contribution element. The path to public-employee pension reform in such states is quite easy: the politics of pension reform may still be tricky, but legally, all the legislature needs to do is amend the statute — even to the point of wiping out the pensions entirely.

Most states have moved away from the gratuity approach. But that doesn’t necessarily mean all public-employee pensions are binding contracts. Some states consider pensions to be more akin to property interests. (And even then, that might imply strong property protection, as in New Mexico, or weak property protection, as in Maine.) Some states consider pensions to be promises of a sort, but ones that are only protected if the beneficiary has reasonably relied on the promise. (That’s the approach in Minnesota.)

And even in the states that consider pensions to be binding contracts, there’s still the question of what’s included in the contract.

In Florida, for instance, the legislature can change pension benefits as much as it wants, as long as it protects the amounts that have been accrued so far. Oklahoma provides less protection: the legislature can change your pension benefits until the moment you retire.

Perhaps the high point of pension protection is in California and states that follow a similar approach, like Arizona and Illinois. In those states, the contract is taken to include not only the amounts earned so far based on work already done, but also the formulas that were part of the law at the moment one was hired (e.g., a 3% annual cost of living increase, a 5% employee contribution rate, etc.). So in those states, even a purely prospective reduction in the annual cost of living increase from 3% to 2%, or an increase in the employee contribution rate from 5% to 6%, would be unconstitutional. A recent California appellate case has seemed to back away from this extremely protective approach, but it remains to be seen whether this move will be upheld by the California Supreme Court. (As noted earlier, Arizona and Illinois have special Pension Clauses, which might explain their more aggressive approach; but California follows a similar approach under its ordinary Contract Clause, though it lacks a Pension Clause. The precise constitutional language probably matters less than the general attitudes of judges in different states.)

How can pension reformers proceed in those “California Rule” states? One possibility is to make pension reforms apply only to new hires. (One’s ability to take advantage of this by firing existing employees will presumably be limited by civil-service laws and government-employee unions.) Another possibility is to amend the state constitution — an approach taken in Arizona, where the Pension Clause was surgically amended to exempt one particular pension reform statute.

* * *

These interlocking questions—Is there a contract? What does it cover?—are nicely illustrated in Kentucky, which has major pension underfunding problems and is now going through a pension reform of its own.

In 1972, the Kentucky legislature passed the following statute:

It is hereby declared that in consideration of the contributions by the members [of the Kentucky Employees Retirement System (KERS)] and in further consideration of benefits received by the state from the member’s employment, KRS 61.510 to 61.705 shall . . . constitute an inviolable contract of the Commonwealth, and the benefits provided therein shall . . . not be subject to reduction or impairment by alteration, amendment, or repeal.

Why would such a statute have been necessary? As noted above, it would make clear that the pension is a contract, not a gratuity. Otherwise, § 3 of the Kentucky Constitution would apply: “every grant of a . . . privilege . . . shall remain subject to revocation, alteration or amendment.” Instead, § 19 applies—a conventional Contract Clause, under which “[n]o . . . law impairing the obligation of contracts[] shall be enacted.” The “inviolable contract” statute serves a useful function — sending pensions covered by KERS into the § 19 box rather than the § 3 box.

So the contract is inviolable. But what’s in the contract—what features can’t be reduced or impaired? Perhaps, like in California, the contract is an agreement to retain pension formulas at least as generous as those that were in effect when the employee was hired. (The explicit reference to particular statutory sections might be taken to endorse that view: perhaps the contract includes the 5% employee contribution rate specified in KRS 61.560, and the specific cost of living adjustments specified in KRS 61.691.) Or perhaps, like in Florida, the contract is just an agreement to pay the pension benefits that have been earned based on work already done. (On this view, the reference to statutory sections just clarifies that these sections apply, however they may be modified by the legislature going forward.)

The statute itself doesn’t say which is the true meaning, and Kentucky caselaw has been fairly sparse. Litigation doesn’t arise unless someone tries to change the system in a way that’s unfavorable to public employees, so states with a long history of pension-reform efforts will naturally have more judicial decisions on the subject. The leading recent case in Kentucky is Jones v. Board of Trustees of Kentucky Retirement Systems (1995). In Jones, the Board of Trustees determined that the state should increase its contributions to the system; the governor and legislature refused to do so. The dispute revolved around differences in how to value retirement system assets — whether to use market value or book value. The Kentucky Supreme Court held that the governor’s and legislature’s actions were legal, on the theory that the Kentucky Constitution and statutes protected actual benefits against “substantial impairment,” and didn’t mandate any particular funding mechanism. Because no substantial impairment had been shown, the Court wrote, “we need not decide under what circumstances the state’s contract with its workers could be lawfully impaired.” So there appears to be some continuing ambiguity here.

In 2013, the legislature amended the statute. Now, the “inviolable contract” blockquote above is prefaced with “For members who begin participating in the Kentucky Employees Retirement System prior to January 1, 2014.” For later hires, a new subsection says:

For members who begin participating in the Kentucky Employees Retirement System on or after January 1, 2014, the General Assembly reserves the right to amend, suspend, or reduce the benefits and rights provided under KRS 61.510 to 61.705 if, in its judgment, the welfare of the Commonwealth so demands, except that the amount of benefits the member has accrued at the time of amendment, suspension, or reduction shall not be affected.

So at least with respect to new hires who participate in KERS, the law seems clear: the legislature is free to change pension benefits, as long as it does so prospectively. As to older hires, the legislature may also have that ability, but there’s more legal uncertainty.

The same is true of some other Kentucky retirement systems, like the Kentucky Legislators’ Retirement Plan or the Judicial Retirement Plan, which have a similar pre-/post-2014 dichotomy.

But some Kentucky plans lack “inviolable contract” language entirely. One recent case, litigated in federal court under Kentucky law, was Puckett v. Lexington-Fayette Urban County Government (2016), where retired policeman Tommy Puckett challenged a legislative reduction of his cost-of-living adjustments. The Sixth Circuit rejected a challenge to this reduction under the federal Contract Clause. Puckett was a member of the Lexington-Fayette Urban County Government Policemen’s and Firefighters’ Retirement Fund, which—unlike KERS and similar funds—wasn’t created by a contract-looking statute. The Sixth Circuit held that because the statute showed no intention of creating a contract (such as “inviolable contract” language), there was no Contract Clause violation.

Admittedly, the Sixth Circuit’s analysis was only under the federal Contract Clause, not the state Contract Clause, and it’s theoretically possible that Kentucky’s Contract Clause might be stricter; but Kentucky caselaw doesn’t strongly suggest that possibility. The fact that Tommy Puckett’s cost-of-living adjustments could be reduced—even years after he had retired, therefore not at all prospectively—suggests that the legal barriers to pension reform are fairly low in Kentucky.

* * *

What does the current pension reform effort look like in Kentucky? The current plan proposed by Gov. Bevin and legislative leaders covers a number of different public pension systems—not just KERS, but also the Teachers’ Retirement System, the County Employees Retirement System, the State Police Retirement System, and several others. The full proposal is big and complicated, and space doesn’t permit a full analysis here.

But let’s look, for instance, at the Teachers’ Retirement System (TRS) component, which is briefly summarized here and somewhat more lengthily summarized here.

The basic idea is to switch employees from the current defined benefit plan into a defined contribution plan. But rather than doing so immediately, the plan adopts a gradual approach. Current teachers can continue to participate in the current defined benefit plan until they reach 27 years of service, or age 60—that is, until they’ve accrued full unreduced retirement eligibility. If they’ve reached that threshold by July 1, 2018, they move into a Social Security replacement defined contribution plan—but they can still opt to continue in their defined benefit plan for up to three more years. If they haven’t reached that threshold by July 1, 2018, they still participate in the defined benefit plan until 27 years or age 60, and then move into the defined contribution plan. Eventually all current employees would be shifted into the defined contribution plan, but the shift would be gradual.

Some of the individual changes are unambiguously negative: for instance, TRS members are currently entitled to a 1.5% cost of living adjustment, whereas the new plan would suspend cost of living adjustments for five years. Whether the net effect of all the changes is positive or negative depends on the interaction of a lot of moving parts, which is beyond the scope of this article.

Is all this legal? It’s hard to say because, as noted above, the contours of the “inviolable contract” for the pre-2014 hires are a bit up in the air. But, broadly speaking, Kentucky law seems to be friendlier to reform efforts than the law of many other states. And even if the contract does include existing formulas, retirees are only protected against substantial impairments that are unreasonable and not backed by a significant public purpose. So if this reform proposal passes, it may be able to avoid the legal hurdles that have bedeviled pension reform efforts elsewhere.

[NOTE: Edited to remove references to Maryland law and insert references to Florida law.]