Five Options for Addressing
‘Transition Costs’ When Closing
the MPSERS Pension Plan

Michigan Public School Employee Retirement Plans
in Need of Reform

This study considers the supposed ‘transition costs’ that would be effected by a state switch from a defined-benefit to defined-contribution retirement system. In it, the “transition costs” are found to be nonbinding and discretionary. In addition, the study offers the state a series of reforms that would diffuse such costs, as well as consideration for the long-term fiscal improvements that would arise from payment of the pension’s unfunded liabilities.

Executive Summary

As of Sept. 30, 2010, the defined-benefit pension plan in the Michigan Public School Employees’ Retirement System had an unfunded liability of $17.6 billion. This unfunded liability and the large annual payments necessary to fund it suggest the plan’s liabilities should be contained by closing the plan to new entrants, much as the defined-benefit pension plan in the Michigan State Employees’ Retirement System was closed in 1997. Future public school employees would be offered participation in a 401(k)-style defined-contribution plan.

A concern repeatedly raised about closing the MPSERS plan, however, is the so-called “transition costs” involved. This paper discusses these “transition costs,” their validity and ways to minimize or eliminate them if they are considered barriers to the important and necessary reform of closing the MPSERS defined-benefit plan.

The Senate Fiscal Agency, the House Fiscal Agency and the Office of Retirement Services have each published estimates of the “transition costs” involved in closing the MPSERS defined-benefit plan. The two material items discussed in the papers are changes in the “normal cost” and the “amortization costs.”

Normal cost refers to the annual cost of paying for employee pension benefits earned during that particular year. According to the SFA, HFA and ORS analyses, these costs would increase if the defined-benefit plan were closed and new entrants were placed in a defined-contribution plan similar to MSERS’. This conclusion is questionable. The stated normal cost for the MPSERS defined-benefit plan has almost certainly been understated by a debatable assumption of 7 percent to 8 percent annual pension asset growth. The failure of this assumption is clear from the plan’s asset growth over the past 14 years and the plan’s large and growing unfunded liabilities.

Even assuming the normal costs are as currently stated, the transition cost could be eliminated by simply tailoring a MPSERS defined-contribution plan differently than the MSERS defined-contribution plan. This same MPSERS normal cost structure could be maintained in a defined-contribution plan simply by requiring school districts and other MPSERS employers to contribute approximately 71 cents for each dollar of employee contributions up to 5.38 percent
of employee salary — a total employee and employer contribution similar to that of the MSERS defined-contribution plan in fiscal 2010.

The larger “transition cost” discussed in the three papers is the immediate increase purportedly required in the size of MPSERS’ “amortization payments,” which are deposits made annually by MPSERS employers to pay down the defined-benefit plan’s unfunded liabilities. These amortization payments ensure the pension plan can meet its future pension obligations.

The rules of the Governmental Accounting Standards Board are frequently interpreted to mean that if the MPSERS defined-benefit plan were closed to new entrants, the MPSERS amortization payment schedule would need to change to a “level-dollar” treatment that would calculate larger upfront amortization payments than those projected if the plan were to remain open. The expected increase is large in the first year — the Office of Retirement Services places it at $360 million — but would decline over the next seven years and ultimately produce lower projected payments than under MPSERS current “level-percentage” amortization schedule. In fact, if all went as expected, there would be a projected net savings in total amortization payments over time.

Three things should be noted about this immediate “transition cost.” First, GASB rules govern how MPSERS and other government entities report their expenditures; the rules do not tell policymakers what expenditures they should make, as GASB itself has confirmed. Thus, even if the state were to adopt a level-dollar amortization payment schedule in reporting on a closed MPSERS plan, it would make increased immediate payments only if it chose to.

Second, the phrase “transition cost” is misleading. MPSERS’ cost is the underlying pension liability, which would not change; the cost is not the amortization payments used to meet that liability. Hence, even if MPSERS employers paid more at first, the plan’s cost would not increase any more than a large first payment on a mortgage would affect the value of a home.

Third, the emphasis on normal and amortization “transition costs” ignores a considerable additional cost: the potential that unfunded liabilities in the MPSERS plan will become even more burdensome. In fiscal 2010 (excluding the cost of an early retirement incentive), the unfunded liability in the MPSERS defined-benefit plan increased by $4.3 billion — nearly a tenth of the annual state budget. Since fiscal 2000, the unfunded liabilities of the plan have increased by 6,500 percent, and the state has failed to make GASB’s “required” annual payments eight times.

If policymakers nevertheless perceive “transition costs” for amortization payments as a problem, they have alternatives. Five are provided in the study, listed from most comprehensive to most limited.

First, since MPSERS pension liabilities must be met under the Michigan Constitution, policymakers could scale back MPSERS’ extensive retiree medical care coverage, which is not protected under the Michigan Constitution, and which cost MPSERS employers $795 million in fiscal 2011. The retiree medical reduction need not be $360 million, given that the “transition cost” would decrease in subsequent years, and given that the Legislature could tap $133 million already slated for MPSERS reform.

Second, policymakers should consider making the most of MPSERS reform. If there are upfront “transition costs,” the Legislature could maximize the benefit of incurring those costs by not just closing MPSERS, but also freezing it — that is, providing current defined-benefit plan members with a defined-contribution plan in lieu of their current plan, so that they would earn no additional benefits under the defined-benefit plan. This step would considerably decrease the potential future unfunded liabilities of the MPSERS pension plan. In addition, GASB appears to provide flexibility in choosing an accounting treatment for the amortization payments on a closed and frozen plan, making it even easier for the state to adopt a payment schedule that minimizes immediate costs.

The state’s third-best course is simply to pay the “transition costs.” MPSERS’ unfunded liabilities have built up over years, and they must be made good at some point. Making larger amortization payments sooner rather than later will not only ensure that retirement assets are available as employees retire, but reduce the projected total burden of amortization payments on taxpayers as well.

Policymakers have additional alternatives, however. Legislators have frequently failed to make the “required” amortization payments under GASB schedules, and they could continue this “business as usual” with a closed plan by reporting under a level-dollar schedule, but making level-percentage payments.

Alternatively, legislators could also choose to spread the amortization payments across the entire MPSERS payroll, including the payroll not just of members of the defined-benefit plan, but of the new defined-contribution plan as well — something the state is already doing with MSERS. The state could then maintain a level-percentage accounting treatment for MPSERS.

Each of these five approaches to amortization payments would make it easier for policymakers to honor their constitutional obligation to pay MPSERS pension benefits.

Introduction

Ensuring that state government offers sustainable retirement benefits will be a challenging task for Michigan policymakers. On Sept. 30, 2010, the state had a massive, unfunded $21.7 billion constitutional obligation to provide retirement income that has already been earned by government employees under state government’s two largest pension plans: the Michigan Public School Employees’ Retirement System and the Michigan State Employees’ Retirement System.[*]

To contain these costs, policymakers have begun transitioning some of the state’s various defined-benefit retirement systems to defined-contribution plans.[†] An exception to this reform, however, has been MPSERS, which is the state government’s largest pension plan. As of Sept. 30, 2010, the MPSERS pension plan had total actuarial accrued liabilities of $60.9 billion and an actuarial value of assets of $43.3 billion, leaving an unfunded actuarial accrued liability of $17.6 billion.[1]

Reforming MPSERS by shifting new employees from a defined-benefit plan to a defined-contribution plan would further benchmark the state’s retirement systems to private-sector norms. Yet the state has also been warned by several analysts that this reform would result in substantial “transition costs” the state cannot afford.[2]

It might seem that the state can neither keep its system nor reform it — or to paraphrase Jefferson, that the state has a wolf by the ear and can neither hold him nor safely let him go.[3]

This paper, however, explains how policymakers can honor the commitments made to MPSERS participants while controlling and even eliminating so-called “transition costs.” 

State statutes set parameters for the retirement benefits of all Michigan governments and government-owned entities, such as state universities and community colleges. The state requires public school districts to offer retirement benefits through MPSERS.[‡],[4]

Most state employees, on the other hand, receive pension benefits through the Michigan State Employees’ Retirement System.[5] There are also separate plans for judges, state police and legislative employees.[6] Local governments can offer pension benefits to employees at their own discretion.[§]

In 1996, the state created a defined-contribution plan for all new hires who entered the MSERS retirement system. The MSERS defined-benefit plan was closed to new participants, though existing participants were free to remain in the defined-benefit plan and continue earning benefits there if they chose.[7] The Legislature instituted a similar transition to defined-contribution plans for new hires in its retirement systems for judges and legislative employees.[8]

In the defined-contribution plans for the three systems, the state employer automatically deposits an amount equal to 4 percent of an employee’s salary into an independent account and then matches an employee’s personal contributions to the account up to 3 percent of the employee’s salary.[9] The individual employee is responsible for determining how much money is placed in his or her account, how this money is invested[¶] and how this money is used upon his or her retirement. The state does not guarantee a particular retirement income or incur liabilities for future payments.

These defined-contribution systems offer three major benefits to employers. First, the plan is “current,” meaning that the costs for retirement are paid in full on an annual basis and the employer does not risk having to contribute more money in the future for unfunded liabilities. Under a defined-contribution plan, costs are incurred immediately and payment of that cost retires the employer obligation completely. In contrast, under a defined-benefit plan, employer payments are only deposits set aside to pay a future liability. These deposits may prove insufficient, so that further deposits become necessary to cover the unfunded liabilities that develop. Unfunded liabilities in a defined-benefit plan may emerge for several reasons: investment returns that are lower than the initial predictions; demographic patterns, such as member longevity, that diverge from initial expectations; or future pay and benefit changes.

Second, the plan is predictable, meaning that the employer’s costs fluctuate only with payroll and with the program’s design parameters, both of which are under the employer’s control. The employer’s contributions to the plan are not subject to factors like problematic investment returns or demographics — the less predictable elements that influence employer pension contributions in defined-benefit plans.

Third, benefits in a defined-contribution system are usually affordable. Private-sector employers’ payments to defined-contribution plans are usually between 5 percent to 7 percent of payroll, as noted by actuary and Mackinac Center Adjunct Scholar Richard C. Dreyfuss.[10] MSERS defined-contribution plan, which requires state government deposits of up to 7 percent of employee payroll, is consistent with this general private-sector range.

In contrast, current employer contributions to the MPSERS defined-benefit pension plan are 17.39 percent or 18.62 percent of payroll, depending on when the employee was hired.[11] This high percentage is largely because of unfunded liabilities, which in turn are largely due to experience that has not matched expectations and to the state’s failure to make the actuaries’ annual recommended contributions to the MPSERS pension fund. As noted earlier, at the end of fiscal 2010, MPSERS defined-benefit pension plan had a $17.6 billion unfunded liability.[**],[12]

This unfunded liability suggests that the Legislature should consider repeating its successful MSERS reform with MPSERS, transitioning all new school hires to a defined-contribution plan. In this transition, the MPSERS defined-benefit plan would be closed to new participants, so that they would not increase the plan’s total liabilities and the potential for unfunded liabilities. The MPSERS defined-benefit plan would stop incurring further liabilities with each new hire, making it easier for legislators to finance the unfunded liabilities owed to current employees.

And indeed, as discussed below, the current liabilities in the MPSERS defined-benefit plan are owed to employees under the Michigan Constitution. This does not mean, however, that reforms cannot legally be made to retirement benefits. In fact, such reforms can ensure that the constitutional promises are kept.

Legal Obligations

Public-sector pension reforms are subject to legal constraints. Defined-benefit pensions in government plans are protected by Article 9, Sec. 24, of the Michigan Constitution, which states, “The accrued financial benefits of each pension plan and retirement system of the state and its political subdivisions shall be a contractual obligation thereof which shall not be diminished or impaired thereby.”[13] While this statement seems relatively clear in theory, there has been litigation to determine what it means in practice.

In Studier v. Michigan Public School Employees’ Retirement Board, the Michigan Supreme Court confirmed that employees are entitled to the pensions that they have already earned through participation under the system’s rules so far, but the court did not hold that employees are entitled to future participation in that same system.[*], [14] Going forward, then, a government can “freeze” the pension benefits that an employee has earned, meaning they no longer earn further benefits under their current terms.

This does not mean the state is entirely off the hook, however. If a gap opens between the benefits earned by those employees before the freeze and the money set aside and invested to pay for those benefits, the state cannot renege on providing the earned benefits; employees still have the legal and constitutional right to expect that their earned benefits will be paid. The employees’ right to these benefits is ultimately a claim on the Michigan taxpayer.

In reference to public pension plans, the Michigan Constitution also states, “Financial benefits arising on account of service rendered in each fiscal year shall be funded during that year. …” The Michigan Supreme Court recognizes this constitutional injunction to mean that the state should be “prefunding” these benefits — that is, that pension liabilities should be saved for during the year they are earned, rather than being paid only when they come due years later.[15]

This prefunding is usually accomplished by making annual payments to a pension fund that is invested in order to grow over time (the investment portfolio is managed by the state Department of Treasury[16]). For example, a single year of service might be worth an additional $600 per year in pension benefits for an employee upon retirement. Under a series of assumptions, such as when that employee will retire, how long he or she will collect pension benefits after retirement, and how much the state’s pension fund investments should grow, state actuaries calculate a “normal cost” of retirement to be set aside annually to cover the $600 benefit to be paid each year to the retired employee as a result of the employee’s earlier year of service.[†] The total annual normal payments deposited to the pension fund during the employee’s preretirement years will generally be less than the total amount of money paid to the employee during his or her retirement years, since the earlier normal-cost deposits are assumed to grow through investment returns during the employee’s working life. This prefunding is supposed to ensure that the state does not push its pension costs for current service into the future, and it is thus meant to thwart the temptation for policymakers to provide ever-increasing benefits paid for long after they’ve left office.[‡]

While the state Supreme Court ruled that retirement costs are not supposed to be deferred under the Michigan Constitution, the court also ruled that it does not have the power to prevent the governor and Legislature from deferring them.[17] On the whole, while the state constitution clearly requires the state to pay retirees the pension benefits they have accrued during their working years, the court does not have the power to mandate that these benefits be prefunded.

When pension benefits are not properly prefunded, unfunded future liabilities accrue.[§] Standard pension funding policies encourage the state to make annual payments to completely eliminate its unfunded liabilities over time. This process ensures that sufficient pension funds are available when employees retire. Based on the most recent annual actuarial valuation, school districts were paying down the defined-benefit plan’s unfunded liability over a 26-year amortization period.[18]

The Michigan Constitution states that these unfunded liabilities cannot be paid off, however, using money meant to prefund benefits being earned in any particular year.[19] In accounting terms, then, the constitution requires that the normal cost be paid before any money is used to address unfunded liabilities.

The state constitution is otherwise silent, however, about paying down unfunded liabilities.[¶] It does not require that a particular payment method for unfunded liabilities be followed.

Inevitably, there is more to financing a pension system than the state constitution discusses. The constitution does provide a framework, however, within which the Legislature must operate when considering reforms.

MPSERS and ‘Transition Costs’

Concerns have been raised about the immediate costs of closing the MPSERS defined-benefit plan to new hires and placing those hires in a defined-contribution plan. These costs have been discussed in a Michigan Senate Fiscal Agency paper titled “Examining a Change from Defined Benefit to Defined Contribution for the Michigan Public School Employees’ Retirement System” and in a House Fiscal Agency memo titled “Converting MPSERS from a Defined Benefit (DB) to a Defined Contribution (DC) System.”[20] The costs were also discussed more recently in a paper by the Office of Retirement Services.[21]

The SFA Analysis

The SFA and HFA analyses present similar findings. The SFA report includes more detail, so it is the focus of the discussion below.

The SFA analysis assumes that the MPSERS defined-benefit plan would be closed to new hires, and that these new hires would be placed in a defined-contribution plan for public school employees similar to the Michigan State Employees’ Retirement System’s defined-contribution plan. As noted above, the MSERS defined-contribution system caps the employer’s annual retirement contributions at 7 percent of the employee’s payroll — an automatic contribution equivalent to 4 percent of payroll, with an employer match of any employee contributions up to 3 percent of payroll.

In a defined-contribution plan, such as MSERS’, the employer’s contributions are commonly expressed as a percentage of payroll and can therefore be compared to a defined-benefit plan’s normal cost and its payments toward unfunded liabilities, both of which are also expressed as percentages of payroll. As explained above, the annual normal cost of the pension is calculated by state actuaries and represents an estimate of the amount of money required to prefund the retirement benefits earned by employees in a given year, so that on the whole, if all goes well, their pension benefits are fully funded at the time they retire. Amortization payments toward unfunded liabilities are likewise calculated by state actuaries according to yearly payment schedules determined by a number of assumptions.

The 2009 SFA paper, reflecting on the state’s funding policies and actuarial assumptions, observed that aside from the relatively small costs to develop and administer a new defined-contribution MPSERS retirement system, such a system would cost more than the current MPSERS defined-benefit system in two ways.

• Normal Cost. The public school system’s annual payments to a defined-contribution plan would be higher than the current normal cost of the MPSERS defined-benefit plan, the paper observed.[22] Using the MSERS system as a basis of comparison, the SFA paper found that the normal cost of the existing MPSERS defined-benefit plan is lower: 4.21 percent of payroll, rather than the 6.55 percent of payroll paid by the state to the MSERS defined-contribution system.[*] After adding a supplemental piece for pension benefits offered to defined-contribution members in special cases,[†] a defined-contribution retirement system would cost the state an additional $7 million in the first year, and the extra costs would increase significantly as more members became participants in the new plan.


• Initial Payments on Unfunded Liabilities.
  The initial annual payments toward the unfunded liability in the MPSERS defined-benefit pension plan would rise once the plan is closed to new members, the paper stated.[23] The increase would be due to the shift in accounting treatment that occurs when closing a defined-benefit plan.
  
  When paying down any existing unfunded liability in an open pension system — that is, a pension system accepting new entrants — government accounting rules are generally interpreted to mean that employers should calculate their contributions as a level percentage of the payroll.[‡] As a result, the amortization payments on the unfunded pension liability of an open system are “backloaded” — as payroll increases over time, payments also increase, since these are calculated as a fixed percentage of payroll. This dynamic allows the amortization cost to be a lower percentage of payroll in the early years than it would be if it were paid in equal annual dollar amounts. The payment dynamic is reminiscent of a variable rate mortgage with a rising interest rate, where payments increase in the later years.
  
  When closing a defined-benefit pension system, however, government accounting rules are usually interpreted to require that employers pay for the unfunded liability as a level-dollar amount each year, rather than as a percentage of an eventually declining payroll.[§] In comparison to the level-percentage payment schedule, this level-dollar method is “frontloaded” (though it is technically flat), generating a higher cash payment schedule in the early years. The SFA paper indicates that this level-dollar payment method for a closed MPSERS defined-benefit system would cost $208 million more in payments during the first full year in which the MPSERS system was closed.[24]

The second of these issues — the immediate increase in the calculated payments on the plan’s unfunded liability — is the primary concern over closing the MPSERS defined-benefit pension plan to new public school employees and placing them in a defined-contribution plan. Indeed, this issue was raised even during Michigan’s 1996 state employee pension reforms, when the state closed its state employee defined-benefit plan, but not its school employee defined-benefit plan. According to a Dec. 5, 1996, article from Gongwer News Service, “Public school employees were given a late-hour reprieve from a similar [defined-contribution] plan when administration officials agreed to a compromise proposal that allows [school employees] to remain under the defined benefits system until the state pays off a $6 billion unfunded liability.”[25]

It should be noted that the SFA’s $208 million first-year estimate for the increase in unfunded liability payments is probably no longer accurate. MPSERS unfunded liability has risen substantially since the report was written — from $8.9 billion in fiscal 2008 to $17.6 billion in fiscal 2010. This hike has increased the amortization payments and the likely gap between level-dollar and level-percentage accounting treatments.[26]

The Office of Retirement Services’ Estimates

A recent unsigned, undated, unpublished memo from the Office of Retirement Services also explores costs from transitioning MPSERS to a defined-contribution retirement system. The paper reiterates the three types of costs listed by the SFA paper. These are:

• The calculation that MSERS employer automatic- and matching-contributions exceed the MPSERS employer normal cost
• Increased upfront cash requirements due to actuarial assumption changes
• Relatively small administrative costs from setting up a new retirement system.[27]

There are a few key differences between this memo and the SFA report. There are slight differences in the annual employer service costs. Under the assumption that employees will mimic the contribution rates of employees in MSERS, the ORS paper reports that the defined-contribution plan would cost an extra $9 million in the first year, $20 million in the second, and $32 million in the third.[28] These annual costs increase as more employees enter the defined-contribution system after the defined-benefit system is closed. The ORS’ reported costs are $2 million more, $4 million less, and $6 million less than the figures in the SFA paper, respectively.[29]

Under the assumption that the employees would maximize the employer’s matching contributions, the ORS memo reports that the state’s costs would be $12 million more in the first year, $27 million in the second year, and $42 million in the third year.[30] These costs are $3 million more, $1 million less, and $3 million less than the SFA paper, respectively.[31]

Because the ORS paper is more recent and unfunded liabilities have grown since the SFA published its analysis in 2009, the ORS reports an increase in the cost of conforming to a closed system’s level-dollar schedule to catch up on MPSERS’ unfunded liabilities. The total first year costs are $360 million in the ORS report[32] — $152 million higher than the SFA paper.[33]

The SFA and ORS reports provide similar estimates on the costs of setting up a defined-contribution retirement plan; the numbers, at most $8 million to $10 million, are relatively small in comparison to the calculated increase in cost for the unfunded liabilities.[34]

Dealing With ‘Transition Costs’

Closing the Michigan Public School Employees’ Retirement System to new employees would prevent state legislators from passing retirement system costs on to future taxpayers through unfunded liabilities and questionable tinkering with abstract funding assumptions. There are also ways a defined-contribution reform can minimize or even eliminate the transition costs discussed above.

These dynamic approaches to transition costs appear immediately below. Nevertheless, three observations should be made to put the entire “transitions costs” argument into context, since “high transition cost” is not as definitive an objection as it sounds.

First, using the flat (or relatively “frontloaded”) level-dollar method for paying down unfunded liabilities is (at best) a guideline of the Governmental Accounting Standards Board, not of the Michigan Constitution.[*] The Michigan Legislature is free to amortize its unfunded liabilities as it sees fit, as long as it (or the various state employers) first makes sure to prefinance retirement benefits — that is, to pay the normal cost.[†]

Indeed, GASB rules address how to account for financial problems, not how to solve them. The rules are not meant to guide policy, nor are they meant to allow governments to play games with payment schedules. In fact, the board has emphasized that its rules should not be “about acceptable public policy with regard to an employer’s contributions to a pension plan.”35

Thus, while GASB’s methods of calculating an annual required contribution have been widely adopted by state and local governments, and they will continue to be used in the state’s financial reporting, there are no requirements that they guide funding policy. Indeed, as discussed later, Michigan has frequently chosen to make payments that diverge from them. This history makes the rigid insistence that funding be driven by a different GASB accounting treatment for a closed pension plan seem artificial. GASB’s rules for reporting financial activity should not be a barrier to implementing a defined-contribution policy that would ensure that retirement benefits for future members are affordable, predictable and current.  

Second, in an important sense, the phrase “transition costs” is misleading when applied to the amortization payments on the unfunded liability. If MPSERS pension system is closed, the frontloaded amortization payments recommended under GASB guidelines in order to catch up on unfunded pension liabilities do not represent increased total expenses, only increased immediate cash outlays. The larger upfront deposits are similar to making early payments on a mortgage. The payments do not change the underlying value of the home (though it would lower the total cash payments on the mortgage).

Third, the “transition costs” objection to changing MPSERS to a defined-contribution plan rests on the assumption that the normal cost and the amortization payments on the unfunded liability provide a proper accounting of the costs involved — an incorrect assumption that will be analyzed later in this paper.

Normal Cost

One assumption in the SFA, HFA and ORS analyses will be addressed directly, however: the idea that the current normal cost is an adequate reflection of the cost of prefunding employees’ benefits as they are earned each year. Currently, the normal cost — and particularly the employer’s normal cost — for the MPSERS defined-benefit plan is artificially lowered by the state’s optimistic assumption that nearly all of the plan’s assets will achieve an 8 percent annual rate of return.[*] But from the end of fiscal 1997 to the end of fiscal 2011 — a 14-year period that includes the substantial market returns of the late 1990s — MPSERS’ defined-benefit portfolio realized an approximately 5.4 percent annual rate of return.[36] This performance raises doubts about the reliability of the 8 percent assumption.

Assuming an 8 percent rate of return, however, has reduced the plan’s apparent normal cost (and, of course, led to additional unfunded liabilities). It also means the SFA, HFA and ORS analyses, which implicitly relied on this assumption, have almost certainly overstated the difference between the normal cost of the MPSERS defined-benefit plan and the annual employer deposits to the MSERS defined-contribution plan.

But even taking the SFA, HFA and ORS comparisons at face value, the transition costs due to the “lower” normal cost of the MPSERS defined-benefit plan can be addressed without increasing the employer’s annual obligations. There are no requirements to duplicate the MSERS defined-contribution provisions in every regard.[†] An employer can choose the cost of benefits in a defined-contribution system.

For instance, if the state simply wants to duplicate its employer costs for pension contributions, it can structure a defined-contribution plan accordingly. The state’s most recent actuarial valuation, with figures more recent than those in the SFA analysis, shows that MPSERS’ active members pay an average of 5.38 percentage points of the total 9.22 percent normal cost.[37] This leaves school districts paying only 3.84 percentage points of the normal cost. To maintain the same employer-employee contribution ratio and overall costs, the state would simply require school districts to offer 71 cents for every dollar contributed by the public school employee and cap the schools’ payments at an employee contribution of 5.38 percent of salary.[‡]

Note that the figures above involve average contributions from employers and employees for a single year (fiscal 2010). Individual employees contribute varying percentages based on when they were hired and their annual salaries. The overall normal cost contribution also varies from year to year based on new actuarial assessments of the plan; for example, the figure was 9.55 percent in fiscal 2009.[38]

In a defined-contribution plan, however, the Legislature can choose a particular employer contribution rate, and legislators should recognize that the plan described here is in fact similar in size to the MSERS defined-contribution plan. The money currently set aside for the MPSERS defined-benefit plan is the sum of a 5.38 percent employee contribution and a 3.84 percent employer contribution — a total of 9.22 percent of payroll. The MSERS defined-contribution plan, on the other hand, is a 6.55 percent employer contribution and a 2.55 percent employee contribution — a total estimated 9.10 percent of payroll.[§] In other words, these two plans would set aside similar payroll percentages to pay for retirement benefits.

The remaining question, then, is to address the change in unfunded liability payment assumptions that accompanies the closing of a defined-benefit pension plan.

Amortization Payments on Unfunded Liability

As noted above, the “transition cost” causing policymakers the most concern is that for switching to a level-dollar reporting schedule for paying down the unfunded liabilities of a closed MPSERS defined-benefit plan. The ORS calculates this figure could be $360 million in the first year of the transition (assuming the state made the full payment on a level-dollar schedule), though the figure would be lower and eventually negative in subsequent years, yielding “savings” (see Graphic 1).

Graphic 1: Office of Retirement Services’ Estimates of Unfunded Liability ‘Transition Costs’ in Closing MPSERS Defined-Benefit Plan

Source: “The Cost of Converting the Michigan Public School Employees Retirement System to a Defined Contribution Plan,” Michigan Office of Retirement Services, Page One.

Below are five options for dealing with “transition costs.” If legislators would like to ensure that closing the system will not require additional cash in the near-term, they can consider all of the options below with the exception of option three. In option three, reasons are presented for why legislators could simply embrace these “transition costs” and make payments according to the level-dollar schedule. The five options are presented in order from the most comprehensive to the most limited reforms — and hence presented in the order in which the author recommends them.

There is a special consideration for policymakers interested in closing the MPSERS defined-benefit system in the near future. The fiscal 2012 state budget includes a special $133 million appropriation for the “MPSERS retirement obligation reform reserve fund.” A workgroup has been called to identify reforms to the pension fund and estimate the costs and savings of those reforms.[39] The money set aside could be used to reduce the size of the transition costs independent of the methods discussed below.

1) Reining in “OPEB” and Setting Priorities

While underfunded pensions are a problem, they pale in comparison to the underfunding of other post-employment benefits. These benefits are “comprehensive group medical, prescription drug, hearing, dental and vision coverages for retirees and beneficiaries,” according to the MPSERS annual actuarial valuation report for health benefits in 2010.[40] Michigan provides health insurance coverage to MPSERS retirees in an amount that depends on the years of service they earned and when they were hired. At the maximum, the state pays 90 percent of the cost of the premium.[41]

With these “other post-employment benefits” — frequently called “OPEB” — the state has not set money aside as the benefits are earned, as it has tried to do with pension benefits. Instead, it assesses additional contributions from MPSERS employers to pay for the bills of retirees as they come due — a “pay-as-you go” approach, rather than prepayment.[42] In the Michigan Constitution, there is no requirement to prefund OPEB, as there is for pensions.

Under the state’s current policies, state actuaries project that the present value of all future payments for benefits already earned under the state’s current policies exceed the present value of the few state assets set aside for these benefits by between $16.7 billion and $27.6 billion.[*], [43] The magnitude of this gap is similar to the state’s $17.6 billion unfunded liability for MPSERS defined-benefit pension plan.[44]

To cover the pay-as-you-go costs of providing other post-employment benefits in fiscal 2012 and fiscal 2013, the state is requiring school districts to pay to MPSERS amounts equivalent to 8.5 percent of payroll and 8.75 percent of payroll, respectively. In fiscal 2011, the state set aside $958.8 million to cover this benefit.[45]

The state has latitude to change this benefit for current members and retirees alike. Unlike public employee pension benefits, these retirement benefits are not protected by the Michigan Constitution.[†] Moreover, retiree medical care is a benefit that is rarely afforded in the private-sector. In a 2010 Mackinac Center Policy Brief, actuary Richard C. Dreyfuss found that only three of a sample of 24 major Michigan private-sector employers provided this benefit to members.[46] A key reason such benefits are uncommon is that once retirees get to age 65, they are entitled to Medicare benefits that cover a substantial portion of their health expenses.[‡]

Instead of levying substantial payments on schools and taxpayers for benefits that few private-sector taxpayers can afford, and rather than committing taxpayers and school districts to substantial long-term risk and financial expense, policymakers could scale back this benefit. The Legislature can use the savings to catch up on financing earned pension benefits, which, unlike OPEB, cannot be reduced under the Michigan Constitution.

The level of savings from this would depend on how much health care expenses increase in the future, how much legislators cut back this benefit and whether the reforms are applied to current retirees or to future ones only. In fiscal 2011, MPSERS employers paid $795 million to retiree medical care benefits. The first-year “transition cost” would be $360 million — nearly half of the 2011 retiree medical coverage payments.

This figure does not mean legislators would need to cut retiree medical costs in half, since reductions to retiree medical coverage would not have to net $360 million in savings each year. The “transition cost” for the amortization payments is lower in subsequent years, and the Legislature has already set aside $133 million that can be used to mitigate the upfront “transition costs.”

It is worth noting that the state has scaled retiree medical benefits back in the past. The state's 1996 reforms revised MSERS retiree benefits to require 30 years of service before an employee received 90 percent of a retiree’s health care premium, with employees earning 3 percent of the health care premium for each year of service. Previously, an employee was fully vested in retiree health benefits after just 10 years of service.[47]

2) Freeze and Close: Maximize the Results

The reason policymakers may close the MPSERS defined-benefit system is to provide some containment of growing unfunded liabilities. The concern over transition costs and potential political conflict also suggests that the remedy should be worth the trouble. In that case, legislators could also consider not just closing the defined-benefit plan to new members, but freezing the benefits earned by current members.

Consider, for instance, how lopsided employer contribution rates are for new MPSERS members. As Graphic 2 shows, in the coming fiscal year, hiring new public school employees in Michigan will require a school system to pay not only their salary and benefits, but also an additional contribution — more than 26 percent of their salary — into MPSERS for OPEB and pension benefits.[*], [48] Only 2.24 percentage points of those contributions pay for the pension benefits that the employees will earn for their work that year — i.e., the employer’s portion of the pension’s normal cost.[49] Much of the remainder — 12.49 percentage points — will pay off previous unfunded accrued liability.[†] For new employees, then, a substantial portion of the retirement money that school districts set aside after hiring them actually will pay for benefits that other employees and retirees have already earned.

As discussed under “Normal Cost” above, one reason for the low employer normal cost is the state’s questionably optimistic assumptions about the plan investments’ rate of return. But another reason for the relatively low employer normal cost is the state’s mandate that most members contribute directly to MPSERS by making payments straight from their paychecks toward their retirement benefits.[‡] In fiscal 2010, active members hired before July 1, 2010, contributed an average of 5.42 percent of their salary to the system, while active members hired on or after that date contributed 4.69 percent.[§],[50] Most MPSERS members are required to make contributions on a sliding scale, ranging up to “$510.00 plus 6.4 percent of the excess [of salary] over $15,000.”[¶],[51]

Graphic 2: MPSERS Employer Contribution for Fiscal 2013 as a Percentage of Payroll

Source: “Employer Contribution Rate,” (Michigan Office of Retirement Services, 2012), http://goo.gl/xDLlk (accessed March 4, 2012).

Freezing the plan means the state would honor all pension benefits already earned and all liabilities already generated, but would no longer allow existing active MPSERS members to earn new benefits (and generate new state liabilities) in the defined-benefit system. This step would also likely eliminate all of the employee payments to the MPSERS defined-benefit system, since these employee contributions are currently dedicated to paying the normal cost of any new pension benefits being earned, and under a freeze, no new benefits would be available. As discussed above under “Normal Cost,” this money would then be available as contributions to a new MPSERS defined-contribution plan.[**]

GASB guidelines are not entirely clear about the preferred accounting treatment for amortization payments on the unfunded liability in a frozen-and-closed system. In light of this, the state could consider constructing a responsible but backloaded amortization schedule. The state would meet its constitutional requirement to honor the accrued financial benefits of MPSERS retirees. Ideally, as discussed under “3) Just Pay It,” MPSERS employers would shorten the amortization period to pay these liabilities down within the average working lifetime of the workers who earned benefits under the defined-benefit plan.

Alternatively, policymakers could consider continuing to use a level-percentage payment method, since the only caveat GASB states explicitly for the use of a level-percentage payment method with a closed plan is that the actuaries assume a decrease in membership over time.[52] The level-percentage payment in a frozen-and-closed system would no doubt be higher than in an open system, but upfront “transition costs” might be mitigated by payroll increase assumptions.

Finally, if policymakers still decided to follow a level-dollar payment schedule and faced upfront “transition costs,”[††] they would at the same time be adopting a reform that created an even higher long-term cost containment than simply closing the plan would. MPSERS defined-benefit plan would incur no new liabilities for further service, even for current public school employees.

Freezing-and-closing would relieve uncertainty about Michigan’s public pensions by producing a predictable and a well-funded pension system. While the state might still face upfront “transition costs,” depending on the payment it adopts, this reform would fortify the protections afforded to taxpayers. Moreover, this reform could be coupled with others, such as reducing OPEB, to address any remaining “transition costs.”

3) Just Pay It

As mentioned above, the phrase “transition costs” is misleading when applied to the unfunded liability, since the underlying cost of pension benefits does not change as a result of closing the plan. The shift in accounting treatment to a level-dollar schedule is a matter of the timing of the payments of that cost.

Moreover, if the state met the level-dollar payment schedule, it would actually lower the net cash necessary to pay down the unfunded liability. The earlier payments would essentially allow deposits into the plan to grow over longer periods of time. This, in turn, would allow the total deposits over time to be smaller than they would be under a backloaded level-percentage payment schedule. 

The state’s current timing of payments is questionable. The amortization period for the net unfunded liability in the MPSERS defined-benefit plan was stated as 26 years in the 2010 annual actuarial valuation.[53] Since the average active plan member was 45.2 years old in fiscal 2010, this amortization period far exceeds the likely remaining working life of the average MPSERS plan member.[54] The better policy would be to ensure that the benefits are fully paid up — i.e., that the unfunded liability is completely paid down — as workers retire. This would imply a much shorter period for the amortization window — say, 15 years — and require more cash regardless of whether the payment schedule involves a level dollar or a level percentage.

In that sense, the state could consider embracing the level-dollar funding policy of a closed defined-benefit plan and simply paying more cash up front. In fact, it’s not clear the state should be using a level-percentage funding policy in the first place. The number of MPSERS active members and their corresponding payroll has been decreasing, rather than increasing, of late. According to the SFA, payroll hit a peak of $10.4 billion in fiscal 2004, but declined to $9.9 billion in fiscal 2009, the most recent year for which SFA provided data.[55] This decreasing payroll and number of active members simulates the dynamics of a closed plan and would presumably call for the more conservative treatment of a level-dollar reporting schedule.

The state already has some precedent for devoting more cash to liabilities of the pension system. The 2012 budget includes $280 million in extra money from the state’s general fund to begin prefunding MSERS retiree health care costs.[56]

The state has also recognized that more cash may be necessary to prefund MPSERS retirement benefits. In the past, the state has assumed an 8 percent annual rate of return on the investment of MPSERS’ assets, but Public Act 75 of 2010 revises the assumed rate of return to 7 percent for the assets of members who joined MPSERS on or after July 1, 2010.[57] This lower rate assumption increases the amount of upfront cash required to prefund MPSERS pension benefits.

Regardless of whether the state pays its unfunded liabilities using a backloaded or a flat (and therefore relatively frontloaded) method, the payments do not represent an increased expense for the underlying benefits paid to participants. Hence, this discussion of level-percentage and level-dollar payments is entirely one of timing. The state developed its unfunded liabilities by deferring the costs of compensation into the future, so the difference has to be made up at some point. Doing it sooner rather than later is justifiable, even though it means that the state Legislature has to forgo other spending in the near term. In other words, simply abiding by the change to level-dollar payments may be appropriate.

4) Business as Usual: Do Not Pay the Full “ARC”

While the state should consider embracing the switch to a level-dollar payment schedule, policymakers could also note that in the past, the Legislature has not met the annual required contribution computed under GASB guidelines.

For instance, in 2007, Michigan legislators voted to skip the calculations for paying down MPSERS’ defined-benefit plan’s unfunded liability and simply pay “4.5 percent of the unfunded actuarial accrued liability.”[*] This essentially equated to paying the interest on the debt, but not any of the principal.[58]

Similarly, legislators have twice marked the MPSERS and MSERS defined-benefit plans’ assets to market values since 1997. This revaluation had to be done legislatively, and it resulted in increasing the stated value of MPSERS and MSERS pension fund assets by $4.6 billion and $1.3 billion, respectively, in fiscal 1997, and by $3.1 billion and $779 million, respectively, in fiscal 2006. This bumped the total of the stated asset values of the two defined-benefit plans by 17.8 percent in fiscal 1997 and 7.8 percent in fiscal 2006.[59] These moves were made specifically to temporarily lower pension contributions at the expense of future costs and violated the rationale for the five-year smoothing process, which ensures smaller year-to-year fluctuations in the state’s annual pension contributions.

Despite manipulating the rules to policymaker’s advantage, and even aside from intentional underfunding of the pension system, the state’s funding policies have proved insufficient to ensure payment of the actuarially determined annual required contributions to MPSERS or MSERS anyway (see Graphic 3).[†]

Graphic 3: MPSERS and MSERS Annual Payments and Required Contributions, 2000-2011

Source: “Michigan State Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2009,” (Michigan Office of Retirement Services, 2009), 44, http://goo.gl/LKl6O (accessed Feb. 15, 2012); “Michigan Public School Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2009,” (Michigan Office of Retirement Services, 2009), 47, http://goo.gl/Y7tfQ (accessed Feb. 15, 2012); “Michigan Public School Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2011,” (Michigan Office of Retirement Services, 2011), 47, http://goo.gl/hyUcr (accessed Feb. 15, 2012); “Michigan State Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2011,” (Michigan Office of Retirement Services, 2011), 44, http://goo.gl/0drvN (accessed Feb. 15, 2012).

Graphic 3 shows the ARC and the actual payments for both MPSERS and MSERS since 2001. Note that in most years, the actual payments are less than the ARC. In other words, the “annual required contributions” are not legally required — or perhaps more to the point, missing the ARC does not appear to have material consequences, especially if the “underpayment” is not a gross departure and the normal cost is paid.

Thus, the state has ignored GASB’s implied funding policy when convenient; in some cases, the state has simply failed to pay the ARC. Hence, if the state has departed from paying the ARC in pursuit of questionable policies, it can consider departing from paying the ARC in pursuit of better policies.

Specifically, the Legislature could just decide to adopt the level-dollar payment schedule, but fail to meet the payment schedule, just as it has in the past. The “transition costs” could be zero if the Legislature decided to make them so.

There are potential downsides. The state would note its failure to pay the ARC in its financial statements, and bond raters and buyers could react negatively to this departure from the implicit norm.

Still, small changes to the unfunded liability payment method could well be considered somewhat inconsequential by the bond marketplace if the state is finally pursuing a legitimate defined-contribution strategy to cap and retire its huge unfunded MPSERS pension liabilities. In fact, following the Legislature’s 2007 underpayment and a second departure from standard practice, there was a note in the MPSERS financial statements that remarked on the changes.[60] If there was a notable reaction in the bond market, it was too small to have a material impact on the rest of state policy.

This is not to say there is no value to GASB standards, and there is every reason for the state to abide by those standards in its reports, even if it does not make the ARC. As Andrew Biggs of the American Enterprise Institute has written, “[G]iven governments’ track record of underfunding their pensions I think these rules (however badly designed they currently are) have some purpose.”[61] Nevertheless, GASB is simply suggesting an accounting change. Allowing this to prevent significant and necessary pension reforms is penny-wise and pound-foolish.

5) Apply Amortization Payments to All Employees

Under most pension funding policies, employers calculate payments toward unfunded liabilities for a particular benefit as a percentage of the payroll of the employees who will receive those benefits. This convention can be changed, however. For instance, Michigan began computing employer contributions toward the unfunded liability for the MSERS defined-benefit plan by using a percentage of the payroll of the employees not just in the defined-benefit plan, but in the defined-contribution plan as well, even though the latter will not receive pension benefits from the defined-benefit plan.[1],[62]

This general approach could be applied consistently to other pension funding policies. For instance, the state could keep its current level-percentage amortization policies and apply the percentage to the payroll for both the new defined-contribution plan and the closed defined-benefit plan.

A similar approach was recently taken in Utah, which passed substantial reforms of a number of its government pension systems. State legislators there were similarly concerned about the immediate cost of the level-dollar payment schedule that is typically applied to a closed pension system.

So the Utah Legislature did not close the system. Instead, it maintained the same backloaded, level-percentage method for paying down unfunded liabilities by adding a “tier” for new employees to the existing pension system. Members of the new tier can opt for a defined-contribution pension plan or for a “hybrid” plan that caps the employer contribution at 10 percent of pay[†] and requires employee contributions to cover any actuarially determined cost above that. State employers set aside a percentage of the payroll for all employees — not just employees in the original tier — as payments on the unfunded liabilities of the existing defined-benefit system. In other words, each new employee’s income was included in the payroll calculations for the unfunded liabilities, regardless of his or her choice of plans as part of the new tier.[63]

In Michigan, members placed in a new MPSERS tier would be treated like the participants in MSERS defined-contribution plan following the 2011 MSERS reform. Under that reform, the state began applying amortization payments to participants in the MSERS defined-contribution plan, not just to participants in the (closed) MSERS defined-benefit plan.64 With MPSERS, then, the state would still use the payroll of new employees — that is, participants in the defined-contribution plan — when calculating the payroll percentage that school districts and other MPSERS employers would need to contribute to the unfunded liability in the MPSERS defined-benefit plan.

Under this approach, the state would not have to alter its amortization schedules and would still be able to use its backloaded method for catching up with its unfunded liabilities. Because there would be no switch to the level-dollar schedule typically recommended for a closed system, there would be no “transition costs” associated with paying unfunded liabilities.

The Benefits of Reform

The preceding discussion shows there are several avenues for reforming MPSERS while minimizing or eliminating near-term “transition costs.” Each has costs and benefits, and it is important not to blow the costs out of proportion or to ignore the benefits.

Take the question of “transition costs.” The papers on this issue by the SFA, HFA and ORS analyze the near-term impact of a defined-contribution reform on both the normal cost and unfunded liability payments. The papers do not consider, however, a very real risk: the possibility that unfunded liabilities will become burdensome to present and future generations in a defined-benefit system. Recent history indicates just how much unfunded liabilities can increase: From fiscal 2009 to fiscal 2010 in the MPSERS defined-benefit plan, the unfunded liability grew from $12.0 billion to $17.6 billion.[65] Some of this increase, of course, was due to an early retirement incentive, but at the very least, the unfunded liability due to the nature of the defined-benefit plan was $16.3 billion.[66] This $4.3 billion increase in a single year was equivalent to a tenth of the annual state budget and more than half of the fiscal 2010 general fund.

In fact, since fiscal 2000, the unfunded liability of MPSERS defined-benefit has leapt more than 6,500 percent from a relatively “modest” $246 million. In that same period, the state has failed to make the ARC eight times (see Graphic 3), despite increasing the employee and employer contributions to the plan. All of these costs have hit the state during a time when its economy has been depressed and taxpayers’ personal finances have been stressed. MPSERS employer contribution rates for the defined-benefit pension plan have risen drastically along with the unfunded liabilities, as Graphic 4 shows. 

Graphic 4: Employer Contribution Rates and Unfunded Liabilities for MPSERS Defined-Benefit Pension, 1989-2013

Source: Kerrie Vanden Bosch, email correspondence with Hohman, Sept. 2, 2011; “Michigan Public School Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2011,” (Michigan Office of Retirement Services, 2011), 46, http://goo.gl/hyUcr (accessed Feb. 15, 2012); “Michigan Public School Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2000,” (Michigan Office of Retirement Services, 2000), 31, http://goo.gl/LzHpV (accessed Feb. 15, 2012); “Michigan Public School Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 1997,” (Michigan Office of Retirement Services, 1997), 29, http://goo.gl/9kjvI (accessed Feb. 15, 2012). Note: The Department of Management & Budget letters for MPSERS employer contribution rates in fiscal 1992 and fiscal 1993 are unavailable from the Office of Retirement Services. Unfunded liabilities for the MPSERS defined-benefit plan are not available for fiscal 2011, fiscal 2012 and fiscal 2013. Contributions for MPSERS OPEB are excluded from the data.

In comparison, during this same period, the MSERS defined-contribution plan has accrued no unfunded liability. The required contributions have been made every year, and the employer contribution rates have remained all but unchanged. Indeed, the MSERS defined-contribution plan has been current, predictable and affordable. And unlike the MPSERS defined-benefit plan, the MSERS defined-contribution plan has clearly satisfied the requirements of the Michigan Constitution, having financed all benefits in the year they were earned. In effect, the MSERS defined-contribution plan has been the mirror opposite of the MPSERS defined-benefit plan.

The defined-contribution plan also provides an element of control missing in the defined-benefit plan. The Legislature can halt any employer payments to a defined-contribution plan when cash is tight without generating new, long-term claims on taxpayers.[67] At the same time, note that during Michigan’s decade-long recession, the Legislature has never once reduced contributions to the MSERS defined-contribution plan.

In fact, a defined-contribution program can allow generous contributions on the upper end. A 401(k) plan, for instance, allows a maximum of $17,000 of annual tax-deferred contributions, regardless of whether it is from an employer or employee, though these restrictions are loosened as employees approach retirement age.[68] Given that the average salary of a school employee is $63,024,[69] a maximum employer contribution would represent 27 percent of pay.[*] Note, too, that such generosity would not need to affect decisions about the plan in future years.

In contrast, the Legislature often provides generous increases in benefits that have ramifications for years to come. For instance, with MSERS defined-benefit plan, the Legislature bumped cost-of-living benefits for retirees 6 times in the 16-year period between 1972 and 1987.[70] The effects of those decisions are still being felt in the MSERS plan, even in fiscal 2010, when the MSERS defined-benefit plan recorded a $4.1 billion unfunded liability.[71]

Similarly, the state has generally assumed that MPSERS’ defined-benefit plan investments will achieve an 8 percent return.[†] But as mentioned above under “Normal Cost,” MPSERS’ defined-benefit portfolio realized an approximately 5.4 percent annual rate of return from the end of fiscal 1997 to the end of fiscal 2011.[72] The gap between state investment performance and its estimations adds to its unfunded liabilities while reducing the plan’s apparent normal cost.

The volatile nature of defined-benefit pension expenses is borne by taxpayers. The cost of these plans depends on numerous uncertainties: the life-length of retirees and beneficiaries, how long employees will work before retiring and employees’ final salaries, among other continually changing assumptions. Shifting contribution rates are a risk to state employers and ultimately taxpayers.

Closing the MPSERS defined-benefit plan reduces this risk. Even if that closure prevents the state from backloading its pension contributions, the larger fiscal danger lies in developing unfunded liabilities in the first place and in the number of actions that the state can take to exacerbate that problem, such as failing to make the annual required contribution.

Conclusion

The Legislature has systematically failed to provide the contributions necessary to fully fund the MPSERS defined-benefit pension plan. If the state hasn’t insisted on meeting its payment schedule under current rules, then a temporarily more demanding payment schedule after switching rules should be an immaterial concern. Nevertheless, this paper has offered a number of ways for legislators to mitigate any concerns they may have in closing the MPSERS defined-benefit plan.

The size of the so-called “transition costs” is determined entirely by how legislators choose to close the MPSERS defined-benefit plan. In essence, outside of the relatively small costs for setting up a defined-contribution system, legislators face “transition costs” only if they choose to, as this Policy Brief makes clear. There are viable ways to eliminate, mitigate or accept the additional cash required by GASB rules when closing a defined-benefit system.

State Sen. Dan Liljenquist of Utah, the legislator who spearheaded pension reform in his state, has remarked that pension funds are like chemical spills: They can be long-term problems, take years to clean up and have disastrous consequences. Defined-contribution reforms are necessary for containing this pension “spill,” since in time they stop the problem from spreading.

Though the five reforms discussed above range from comprehensive to limited, each has trade-offs. Two of the ideas — cutting back on OPEB, and freezing and closing the defined-benefit plan — would do more to contain the problem than just closing the system would. Reducing the MPSERS OPEB would scale back a post-employment benefit the state is not obligated to pay under the state constitution and use the proceeds to solidify the pension commitments the state is obligated to pay under the state constitution. 

Freezing and closing the MPSERS defined-benefit plan, on the other hand, would substantially lower the long-term commitments in the pension system, since the state would no longer incur new pension liabilities for current public school employees, not just future public school employees. Even if freezing and closing the plan left residual “transition costs” for paying unfunded liabilities, the additional containment of liabilities in the system would even more substantially limit taxpayer exposure than simply closing the plan would.

Implementing these two ideas, in other words, would include some clean-up efforts as well as containment.

If policymakers feel unable to pursue these, they could simply embrace make the higher upfront payments suggested by a level-dollar amortization schedule. Catching up on unfunded liabilities more quickly means that future taxpayers will not have to contribute as much to cover promises that were made when many of them were too young to vote. A level-dollar payment approach would simply change the time that the liabilities are paid; it would not constitute an additional liability.

Two of the other ideas — not paying the ARC or applying amortization payments to all employees — would still provide containment while avoiding upfront cash contributions. They would not be as comprehensive a reform, but they would deal with much of the opposition to closing the pension system.

MSERS remains a case study of the benefits of transitioning to a defined-contribution plan. While the state has developed additional unfunded liabilities in the MSERS defined-benefit plan since closing it March 31, 1997, a recent estimate by Mackinac Center Adjunct Scholar Richard C. Dreyfuss indicates that closing the plan reduced the unfunded liabilities by between 36 percent and 51 percent from what they would have been otherwise.73

Of course, there were no “transition costs” for amortization payments when the MSERS plan was closed; the plan was fully funded. But the unfunded liabilities in the MPSERS plan do not mean the reform no longer makes sense; rather, they are powerful evidence of the need for reform.

Michigan policymakers have a range of options to deal with the shift in accounting rules that comes when a defined-benefit plan is closed to new entrants. So-called “transition costs” should not prevent the state from following the private-sector’s lead and offering new public school employees a defined-contribution plan that is current, affordable, predictable and constitutional. By closing MPSERS defined-benefit plan, the state can help ensure that it meets the promises it has made to retirees and vested employees while capping the state’s open-ended and expensive obligations in the future.

Endnotes

1  “Michigan Public School Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2011,” (Michigan Office of Retirement Services, 2011), 46, http://goo.gl/hyUcr (accessed Feb. 15, 2012).

2  See, for instance, Kathryn Summers-Coty, “Examining a Change from Defined Benefit to Defined Contribution for the Michigan Public School Employees’ Retirement System,” (Michigan Senate Fiscal Agency, 2009), http://goo.gl/jnzaB (accessed May 8, 2010), and Bethany Wicksall, “Converting MPSERS from a Defined Benefit (DB) to a Defined Contribution (DC) System,” (Michigan House Fiscal Agency, 2009), https://www.house.mi.gov/hfa/PDFs/MPSERS.pdf (accessed Feb. 17, 2012). See also the undated and unposted “The Cost of Converting the Michigan Public School Employees Retirement System to a Defined Contribution Plan,” (Michigan Office of Retirement Services).

3  Thomas Jefferson, letter to John Holmes, April 22, 1820 (Library of Congress), http://goo.gl/WYMsr. See also “Monticello: Wolf by the ears,” (Thomas Jefferson’s Monticello), http://www.monticello.org /site/jefferson/wolf-ears (accessed Jan. 26, 2012).

4  MCL 38.1301 – 38.1467

5  MCL 38.1 – 38.69

6  MCL 38.2101-38.2670; MCL 38.1601-38.1689; MCL 38.100-38.1080.

7  State Employees’ Retirement Act, MCL 38.1i(3) – 38.1i(4), http://goo.gl/2QRzQ (accessed Feb. 13, 2012).

8  Michigan Legislative Retirement System Act, MCL 38.1018, http://goo.gl/cC3Z1 (accessed Feb. 17, 2012); The Judges Retirement Act Of 1992, MCL 38.2651, http://goo.gl/UCgS9 (accessed Feb. 17, 2012).

9  “State of Michigan 401(K) and 457 Plans,” (State of Michigan), 2, http://goo.gl/ygsyU (accessed Feb. 17, 2011).

10  Richard C. Dreyfuss, “Michigan’s Public-Employee Retirement Benefits: Benchmarking and Managing Benefits and Costs,” (Mackinac Center for Public Policy, Oct. 25, 2010), 4, https://www.mackinac.org/archives/2010/S2010-05.pdf (accessed March 28, 2011).

11  “Employer Contribution Rate,” (Michigan Office of Retirement Services, 2012), http://goo.gl/xDLlk (accessed March 4, 2012).

12  “Michigan Public School Employees’ Retirement System 2010 Annual Actuarial Valuation Report,” (Gabriel Roeder Smith & Company, 2011), A-1.

13  Mich Const 1963, Article 9, Section 24: http://goo.gl/eqXKV (accessed Jan. 23, 2012).

14  Studier v Michigan Public School Employees’ Retirement Board, 472 Mich 642 (2005).

15 Shelby Twp. Police and Fire Retirement Board v. Charter Twp. of Shelby, 438 Mich 247 (1991).

16 MCL 38.1326

17 Musselman v. Governor, 448 Mich 503 (1995), and on rehearing, 450 Mich 574 (1996).

18 “Michigan Public School Employees’ Retirement System 2010 Annual Actuarial Valuation Report,” (Gabriel Roeder Smith & Company, 2011), A-1.

19 Mich Const 1963, Article 9, Section 24: http://goo.gl/eqXKV (accessed Jan. 31, 2012).

20 Kathryn Summers-Coty, “Examining a Change from Defined Benefit to Defined Contribution for the Michigan Public School Employees’ Retirement System,” (Michigan Senate Fiscal Agency, 2009), http://goo.gl/jnzaB (accessed May 8, 2010); Bethany Wicksall, “Converting MPSERS from a Defined Benefit (DB) to a Defined Contribution (DC) System,” (Michigan House Fiscal Agency, 2009), https://www.house.mi.gov/hfa/PDFs/MPSERS.pdf (accessed Feb. 17, 2012).

21  “The Cost of Converting the Michigan Public School Employees Retirement System to a Defined Contribution Plan,” (Michigan Office of Retirement Services).

22 Summers-Coty, “Examining a Change from Defined Benefit to Defined Contribution for the Michigan Public School Employees’ Retirement System,” (Michigan Senate Fiscal Agency, 2009), 3, http://goo.gl/jnzaB (accessed May 8, 2010).

23 Ibid, 4.

24 Ibid.

25 “House Passes Retirement Package Amid Partisan Tension,” Gongwer News Service Michigan Report, Dec. 5, 1996.

26 “Michigan Public School Employees’ Retirement System 2008 Annual Actuarial Valuation Report,” (Gabriel Roeder Smith & Company, 2009), A-1; “Michigan Public School Employees’ Retirement System 2010 Annual Actuarial Valuation Report,” (Gabriel Roeder Smith & Company, 2011), A-1.

27 “The Cost of Converting the Michigan Public School Employees Retirement System to a Defined Contribution Plan,” (Michigan Office of Retirement Services).

28 Ibid., 2.

29 Author’s calculations based on “The Cost of Converting the Michigan Public School Employees Retirement System to a Defined Contribution Plan,” (Michigan Office of Retirement Services) and Kathryn Summers-Coty, “Examining a Change from Defined Benefit to Defined Contribution for the Michigan Public School Employees’ Retirement System,” (Michigan Senate Fiscal Agency, 2009), 4, http://goo.gl/jnzaB (accessed May 8, 2010).

30 “The Cost of Converting the Michigan Public School Employees Retirement System to a Defined Contribution Plan,” (Michigan Office of Retirement Services), 2-3.

31 Author’s calculations based on “The Cost of Converting the Michigan Public School Employees Retirement System to a Defined Contribution Plan,” (Michigan Office of Retirement Services) and Kathryn Summers-Coty, “Examining a Change from Defined Benefit to Defined Contribution for the Michigan Public School Employees’ Retirement System,” (Michigan Senate Fiscal Agency, 2009), 4, http://goo.gl/jnzaB (accessed May 8, 2010).

32 “The Cost of Converting the Michigan Public School Employees Retirement System to a Defined Contribution Plan,” (Michigan Office of Retirement Services), 2.

33 Author’s calculations based on “The Cost of Converting the Michigan Public School Employees Retirement System to a Defined Contribution Plan,” (Michigan Office of Retirement Services), 2, and Kathryn Summers-Coty, “Examining a Change from Defined Benefit to Defined Contribution for the Michigan Public School Employees’ Retirement System,” (Michigan Senate Fiscal Agency, 2009), 4, http://goo.gl/jnzaB (accessed May 8, 2010).

34 Kathryn Summers-Coty, “Examining a Change from Defined Benefit to Defined Contribution for the Michigan Public School Employees’ Retirement System,” (Michigan Senate Fiscal Agency, 2009), 5, http://goo.gl/jnzaB (accessed May 8, 2010); “The Cost of Converting the Michigan Public School Employees Retirement System to a Defined Contribution Plan,” (Michigan Office of Retirement Services), 3.

35 “Accounting and Financial Reporting for Pensions; an amendment of GASB Statement No. 27,” (Governmental Accounting Standards Board, 2011), 98, http://goo.gl/PkvUJ (accessed Sept. 23, 2011).

36 Author’s estimate based on “Michigan Public School Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2001,” (Michigan Office of Retirement Services, 2001), 42, http://goo.gl/RvEJ1 (accessed Feb. 15, 2012); “Michigan Public School Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2011,” (Michigan Office of Retirement Services, 2011), 59, http://goo.gl/hyUcr (accessed Feb. 15, 2012); “Michigan Public School Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 1998,” (Michigan Office of Retirement Services, 1998), 38, http://goo.gl/Vo8qn (accessed March 5, 2012).

37 “Michigan Public School Employees’ Retirement System 2010 Annual Actuarial Valuation Report,” (Gabriel Roeder Smith & Company, 2011), A-1.

38 Ibid.

39 The State School Aid Act, Public Act No. 62, (June 21, 2011), Sec. 147b, http://goo.gl/XU62S (accessed Feb. 1, 2012).

40 “Michigan Public School Employees’ Retiree Health Benefits 2010 Annual Actuarial Valuation Report,” (Gabriel Roeder Smith & Company, 2011), C-1, (accessed Aug. 2, 2011).

41 Ibid.

42 “Michigan Public School Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2010,” (Michigan Office of Retirement Services, 2010), 31, http://goo.gl/CZivR (accessed June 14, 2010).

43 “Michigan Public School Employees’ Retiree Health Benefits 2010 Annual Actuarial Valuation Report,” (Gabriel Roeder Smith & Company, 2011), A-2, (accessed Aug. 2, 2011).

44 “Michigan Public School Employees’ Retirement System 2010 Annual Actuarial Valuation Report,” (Gabriel Roeder Smith & Company, 2011), A-1.

45 Author’s calculations based on “Michigan Public School Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2011,” (Michigan Office of Retirement Services, 2011), 23, http://goo.gl/hyUcr (accessed Feb. 15, 2012).

46 Richard C. Dreyfuss, “Michigan’s Public-Employee Retirement Benefits: Benchmarking and Managing Benefits and Costs,” (Mackinac Center for Public Policy, Oct. 25, 2010), 12, https://www.mackinac.org/archives/2010/S2010-05.pdf (accessed March 28, 2011).

47 “Michigan State Employees’ Retiree Health Benefits 2010 Annual Actuarial Valuation Report,” (Gabriel Roeder Smith & Company, 2011), C-1; “Michigan State Employees’ Retirement System 2010 Annual Actuarial Valuation Report,” (Gabriel Roeder Smith & Company, 2011), F-1.

48 “Employer Contribution Rate,” (Michigan Office of Retirement Services, 2012), http://goo.gl/xDLlk (accessed March 4, 2012).

49 Ronald W. Foss, “All Reporting Unit Business and Payroll Personnel of the Michigan Public School Employees Retirement System,” (Department of Technology, Management & Budget, 2011), http://goo.gl/H4n52 (accessed Jan. 23, 2012).

50 “Michigan Public School Employees’ Retirement System 2010 Annual Actuarial Valuation Report,” (Gabriel Roeder Smith & Company, 2011), A-1.

51 MCL 38.1343a(8).

52 Governmental Accounting Standards Board, “Statement No. 27 of the Governmental Accounting Standards Board,” Journal of Accountancy 180, no. 1 (1995): 130.

53 “Michigan Public School Employees’ Retirement System 2010 Annual Actuarial Valuation Report,” (Gabriel Roeder Smith & Company, 2011), A-1.

54 Ibid., D-4.

55 Matthew Grabowski, “Are Public Sector Retirement Benefits Sustainable? An Update on the State-Administered Retirement Systems,” (Michigan Senate Fiscal Agency, 2011), 3, http://goo.gl/x9P9f (accessed Feb. 16, 2012).

56 Kyle I. Jen and Bethany Wicksall, “State Employee Retiree Health Benefit Prefunding,” (Michigan House Fiscal Agency, 2011), http://goo.gl/YdBe6 (accessed Sept. 26, 2011).

57 “Public Act 75 of 2010,”), Sec. 41 (11), https://www.legislature.mi.gov/documents/2009-2010/publicact/pdf/2010-PA-0075.pdf (accessed Aug. 3, 2010).

58 Kirk Sanderson, “H.B. 4530: Floor Analysis,” (Michigan Senate Fiscal Agency, 2007), http://goo.gl/MSw7A (accessed Aug. 12, 2011).

59 Author’s calculations based on “Michigan Public School Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2000,” (Michigan Office of Retirement Services, 2000), 31, http://goo.gl/LzHpV (accessed Feb. 15, 2012); “Michigan State Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2000,” (Michigan Office of Retirement Services, 2000), 29, http://goo.gl/xljq1 (accessed April 20, 2011); “Michigan Public School Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2010,” (Michigan Office of Retirement Services, 2010), 46, http://goo.gl/CZivR (accessed June 14, 2010); “Michigan State Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2010,” (Michigan Office of Retirement Services, 2010), 42, http://goo.gl/1U8oA (accessed March 17, 2011).

60 “Michigan Public School Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2007,” (Michigan Office of Retirement Services, 2007), 35, http://goo.gl/QFNZx (accessed Sept. 26, 2011).

61 Andrew Biggs, e-mail correspondence with James Hohman, July 27, 2011.

62 Kathryn Summers, “H.B. 4701 (S-1) & 4702: Concurrence Summary,” (Michigan Senate Fiscal Agency, 2011), 5, http://goo.gl/v1kvS (accessed Feb. 14, 2012).

63 Utah State Sen. Dan Liljenquist, telephone conversation correspondence with James Hohman, Sept. 13, 2011.

64 Wicksall, “A Summary of House Bills 4701 and 4702 as Enacted,” (Michigan House Fiscal Agency, 2011), 9, http://goo.gl/siUab (accessed March 4, 2012).

65 “Michigan Public School Employees’ Retirement System 2010 Annual Actuarial Valuation Report,” (Gabriel Roeder Smith & Company, 2011), A-1.

66 Ibid.

67 Vishal Apte and Brendan McFarland, “A Look at Defined Contribution Match Reinstatements,”(Towers Watson, 2011), http://www.towerswatson.com/assets/pdf/5641/Oct11DCMatch.pdf (accessed Jan. 25, 2012).

68 “Contribution limits for retirement accounts,” (The Vanguard Group, Inc., 2011), http://vanguard.com/us/insights /taxcenter/contribution-limits?WT.srch=1 (accessed Sept. 23, 2011).

69 “2009-10 Bulletin 1014,” (Michigan Department of Education, 2011), (v), http://goo.gl/RnGsv (accessed Sep. 9, 2011).

70 “Michigan State Employees’ Retirement System 2010 Annual Actuarial Valuation Report,” (Gabriel Roeder Smith & Company, 2011), F-2.

71 Ibid., A-1.

72 Author’s estimate based on “Michigan Public School Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2001,” (Michigan Office of Retirement Services, 2001), 42, http://goo.gl/RvEJ1 (accessed Feb. 15, 2012); “Michigan Public School Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 2011,” (Michigan Office of Retirement Services, 2011), 59, http://goo.gl/hyUcr (accessed Feb. 15, 2012); “Michigan Public School Employees’ Retirement System: Comprehensive Annual Financial Report for the Fiscal Year Ended Sept. 30, 1998,” (Michigan Office of Retirement Services, 1998), 38, http://goo.gl/Vo8qn (accessed March 5, 2012).

73 Author’s calculations based on Dreyfuss, “Estimated Savings From Michigan’s 1997 State Employees Pension Plan Reform,”(Mackinac Center for Public Policy, 2011), 6, https://www.mackinac.org/archives /2011/2011-03PensionFINALweb.pdf (accessed Feb. 16, 2012).