Current template: single.php


“Pension Liability Stress For State Governments”



This chapter explores the role that public pensions play in the growth of state government spending. Most states provide a defined-benefit style pensions, which are not adequately funded or managed by state governments. Poor funding and management have contributed to unfunded pension liabilities reaching nearly $5 trillion in FY 2018. In many states (such as Illinois), these benefits are constitutionally guaranteed and spending on these retirement payments represents a fixed cost. The more state governments spend on pensions, the less funds available for essential government services, infrastructure, or opportunities for tax cuts. Fortunately, states such as Wisconsin, Michigan, Maine and Tennessee offer comprehensive reform solutions to the pension liability stress.



This chapter explores the role that public pensions play in the growth of state government spending. Most states provide a defined-benefit style pensions, which are not adequately funded or managed by state governments. Poor funding and management have contributed to unfunded pension liabilities reaching nearly $5 trillion in FY 2018. In many states (such as Illinois), these benefits are constitutionally guaranteed and spending on these retirement payments represents a fixed cost. The more state governments spend on pensions, the less funds available for essential government services, infrastructure, or opportunities for tax cuts. Fortunately, states such as Wisconsin, Michigan, and Tennessee offer comprehensive reform solutions to the pension liability stress.

The first section discusses the structure of the defined benefit pension plan (the most common type of public pension plan in the U.S.). The second section discuss how government accounting changes have created transparency and provided a more accurate picture of pension liabilities. The third section will show the differences in risk levels between pension assets and liabilities and then argue why lower discount rates are the best measurement for pension liabilities because they reflect a state’s inability to default on pension promises. The fourth section examines public pension liabilities in the United States utilizing data collected by the American Legislative Exchange Council for its annual pension report. The fifth section examines cases of states enacting successful reforms for their respective pension systems. The sixth section concludes with additional recommendations for pension reform.


Public Pensions Structure: Defined Benefit Plans

Most pension plans are issued in the form of a defined benefit. A defined benefit pension is a pension plan where employees (in the case of public pension plans, state workers) and employers (in the case of public pension plans, state governments) contribute funds during the employees’ time at work and a specified amount of monthly retirement income is provided to the employee upon retirement. That retirement payment is typically based on the employee’s salary, years of work, and age.i This formula is determined by how long the retiree has worked in the public sector and their final average compensation at retirement. Generally, the formula resembles something like Equation 1 below:

(1) 𝐴𝑛𝑛𝑢𝑎𝑙 𝑅𝑒𝑡𝑖𝑟𝑒𝑚𝑒𝑛𝑡 𝐵𝑒𝑛𝑒𝑓𝑖𝑡 = 𝐵𝑒𝑛𝑒𝑓𝑖𝑡 𝑀𝑢𝑙𝑡𝑖𝑝𝑙𝑖𝑒𝑟 × 𝑌𝑒𝑎𝑟𝑠 𝑜𝑓 𝑆𝑒𝑟𝑣𝑖𝑐𝑒 × 𝐹𝑖𝑛𝑎𝑙 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑆𝑎𝑙𝑎𝑟𝑦


Divide the annual retirement benefit by 12 and that determines how much in benefits are paid per month.

Generally, the normal cost (the projected cost to pre-fund retirement promises during the years an employee works) is paid for by contributions from both employees and employers. Most pension plans are not fully funded, and the portion of accrued promised pension benefits that are not covered by plan assets are paid for by the employer and taxpayers (unless the plan is cost sharing, where employees cover these payments as well).

Each year, state governments must make an Actuarially Determined Contribution (ADC). The ADC (previously known as the annual required contribution or ARC) is an annual payment that consists of the normal cost and the amortization payment (a catch-up payment for any unfunded liabilities over the past 30 years).ii If a plan is consistently making 100% of its ADC payments, it is better able to adjust to fluctuating variables (i.e. cost of living adjustments and life expectancy) and pay off its unfunded liabilities in a timely manner.

Illinois has the second largest unfunded pension liabilities in the country at $359 billion (only California has greater unfunded liabilities) and the third largest unfunded liabilities per capita at $28,220 per resident (after Connecticut and Alaska). This is, in part, due to Illinois’ pension contributions failing to meet the ADC due to state statutes Public Acts 100-0023 and 100-0340 using a methodology that does not conform with ADC calculation methods set by GASB. Illinois plans always make payments based upon the state statutes and not the ADC.iii The table below has been recreated from the American Legislative Exchange Council annual report on public pensions with permission from the authors. The table highlights ADC payments for the pension plans in the state of Illinois for fiscal year 2018 (June 30, 2017-July 1, 2018).


Plan ADC ADC Paid Percent ADC Paid
Illinois General Assembly Retirement


$32,084,644.00 $21,155,000.00 65.93%
Illinois Judges Retirement System $168,056,916.00 $135,962,000.00 80.90%
Illinois Municipal Retirement Fund $947,568,823.00 $947,568,823.00 100.00%
Illinois State Employees Retirement System $2,739,377,709.00 $1,929,175,044.00 70.42%
Illinois Teachers Retirement System $7,080,756.00 $4,178,650.00 59.01%
Illinois University Retirement System $1,862,033.00 $1,607,880.00 86.35%

Source: Williams, Jonathan, et. al. Unaccountable and Unaffordable 2019. American Legislative Exchange Council. 2020


The one notable exception, the Illinois Municipal Retirement Fund (which uses ARC methodology to determine the required contribution), has the highest funding ratio of Illinois plans (a lowly 48.80%) and still has nearly $43 billion in unfunded liabilities. After years of not making the required contributions, liabilities have piled up, making Illinois’ plans some of the worst funded pension plans in the country with nearly $360 billion in unfunded pension liabilities.iv

Public pension plans have been grossly underfunded and recent changes in pension reporting (as highlighted in the next section) have shown how poor the funding situation is (with a handful of exceptions). Generally, pensions are underfunded due to a combination of four major reasons:

  1. Intentional underfunding: State policymakers have decided not to pay enough into the fund to meet future obligations, creating debt for future generations. As states fail to make necessary contributions but continue to promise the same benefit payouts, the burden of unfunded liabilities is placed on future
  2. Poor management: Overly optimistic investment return goals, open amortization schedules, outdated or unclear actuarial assumptions, politicized pension boards. As will be discussed later, pension plans have drastically increased the risk in their asset portfolios and investment returns have become increasingly volatile. Later sections will discuss how pension board of trustee governance structures relate to investment performance.
  1. Market conditions and volatility: Recessions, long-term decline in interest rates, and pension systems’ vulnerability to unexpected market
  2. Benefit design issues: Plans that allow retirees to double-dip (receive two pensions), spike their pensions (use salary increases and bonuses to increase their final average salary). Recall equation (1) is partially based on the final average salary. In some cases, employees will use raises and bonuses to increase their final average salary, thus increasing their overall annual retirement

Unfunded pension liabilities totaled anywhere between $1.2-$4.9 trillion for fiscal year 2018.v This number varies due to the various discount rates used to estimate the present value of unfunded liabilities. As mentioned above and in the following sections, the discount rate is used to determine the present value of liabilities. As will be recommended and explained, pension plans should use a lower discount rate to determine the value of unfunded liabilities.


Changes in Pension Reporting with GASB 67 and 68

State governments have experienced increased pressure in their balance sheets from growing pension liabilities. This pressure is becoming more apparent with improved financial reporting. The Governmental Accounting Standards Board (GASB) statements 67 and 68 went into effect in FY 2014 and 2015, respectively. These statements focus on how pension plans measure assets and liabilities.

The changes declared in GASB 67 require plan assets to be valued each year so pension trustees cannot engage in “asset smoothing.” Asset smoothing is a process by which pension investment performance is averaged over a five-year period to “smooth out” swings in market performance. As noted by pension scholars Eileen Norcross and Sheila Weinberg, asset smoothing evens out investment swings and provides plan sponsors with predictability in annual contributions, but simultaneously hides the volatility of pension asset Under GASB 67, pension plan officials must provide an actuarial value of assets (AVA) for that given year, putting an end to asset smoothing. This report’s analysis uses the actuarial value of assets (AVA) that is reported in state pension plan actuarial valuations each year.

Under GASB 68, state and local governments were now obligated to report unfunded pension liabilities on state balance sheets as opposed to just the actuarial determined contribution or ADC. Norcross and Weinberg also note under GASB 68, however, state governments can continue a form of asset smoothing. Governments are permitted to defer the recognition of the difference between the return expected on plan assets and the actual return, with this “deferred inflow of resources” occurring over a 5-year period. This is the same as asset smoothing which permits the sponsor to gradually incorporate any changes to the market value of assets that differ from the expected value of assets over time. They note, “The consequences of this practice remain the same [as the consequences of asset smoothing]. Market declines and gains are only gradually recognized, likely increasing the riskiness of sponsor behavior.”vii

The new information required by GASB 67 and 68 is reported in the “Required Supplementary Information” section at the end of each state’s comprehensive annual financial report (CAFR) and in actuarial valuation documents for each pension plan. These notes include breakdown of the ADC, asset valuations and Fiduciary Net Position for all pension plans, how the pension plan discount rate is calculated and information about liability valuations. The net pension liability is shown in equation

(2) below:

(2) 𝑁𝑒𝑡 𝑃𝑒𝑛𝑠𝑖𝑜𝑛 𝐿𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑦 = 𝐴𝑐𝑡𝑢𝑎𝑟𝑖𝑎𝑙 𝑉𝑎𝑙𝑢𝑒 𝑜𝑓 𝐴𝑠𝑠𝑒𝑡𝑠 − 𝐴𝑐𝑡𝑢𝑎𝑟𝑖𝑎𝑙𝑙𝑦 𝐴𝑐𝑐𝑟𝑢𝑒𝑑 𝐿𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑖𝑒𝑠

If the value of the Actuarially Accrued Liabilities is greater than the Actuarial Value of Assets, the Net Pension Liability will show that there are unfunded pension liabilities. Another important measure of the health of a defined benefit pension plan is the plan’s funding ratio. That is expressed in equation 3 below:

The larger the value of liabilities, the lower the funding ratio, and the less “healthy” a defined benefit pension plan. As recommended by the American Academy of Actuaries, plans should strive for 100% funding ratio or greater.viii

Improved reporting and more accurate estimates of state obligations have shed light on the actual value of unfunded pension liabilities. GASB 67 also provided guidance on how liabilities were to be valued. Prior to GASB 67, public pension plans used the expected return on pension assets to assess the value of liabilities. Economists objected to this valuation, stating that legally guaranteed pension promises should be valued with a lower discount rate (as will be described in detail in the next section). Weinberg and Norcross note that GASB 67 attempts to “split the difference” by valuing liabilities that are covered by pension assets with a higher discount rate and unfunded liabilities with a lower discount rate based on the low-risk return on tax-exempt municipal bonds.ix

Many of the changes in assumptions based on actuarial experience studies conducted in 2016 are still in place today (i.e., inflation assumption remains at 2.25%), while other assumptions have changed. For example, some plans have lowered discount rates drastically (such as several Wyoming state pension plans lowering the discount rate from 7.75% to 7.00% in FY 2017), while other plans have incrementally decreased discount rates (such as the California Public Employee Retirement Multiemployer Fund, which gradually decreased its discount rate from 7.50% in FY 2016, to 7.25% in FY 2017, and then at 7.00% in FY 2018). x, xi


Assumptions matters: Rates of Return and Discount Rates

The discount rate is the rate used to determine the present value of liabilities. Although public plans often use the term “discount rate” and “investment rate of return” interchangeably, the two terms refer to two different aspects of a pension plan. Specifically, the investment rate of return is based on a pension plan’s portfolio of investment assets and what those investments will earn. It looks at the risk of the plan’s investment assets.xii

For public pensions, there are different risk levels with pension assets and pension liabilities.

Over the past four decades, pension asset funds have changed from low-risk, fixed income investments (such as U.S. Treasury bonds) to an increasingly volatile portfolio of stocks, bonds, and alternative investments such as office buildings and golf courses.xiii This is the result of lower bond yields, the desire to chase higher returns, and politicians and plan managers using pension funds to advance their own economic development or political agendas — a perfect storm of bad incentives.

The figure below shows the disparity between assumed rates of return (noted by the dotted line) and the actual annual return on investment (noted by the solid line). As pension plans invest in more riskier assets, meeting the assumed rate of return for that year becomes less likely. Some years this pays off and returns exceed expectations while other yeas fall far short of assumed returns.

Average assumed rate of return hovered just below 8% from 2001 to 2007 but steadily declined from 7.99% in 2001 to 7.88% in 2008. That decline continued in 2009 to a lower 7.85% and then steadily declined each year. By 2018, the average assumed rate of return was 7.22%, a 10.7% decline over 18 years. However, as the graph shows, average actual 1-year returns were much more volatile. During years of recession, investment returns averaged losses of -4.91% in 2001, -6.22% in 2002, -9.49% in

2008, and -9.42% in 2009. Throughout the years measured, 9 of the 18 years measured saw average 1- year investment returns below the average assumed return.

While market conditions contributed to lower funding ratios by lowering the value of assets, the pension crisis is primarily a spending problem that stems from the intentional underfunding of a pension plan and the poor management of the plan. The ALEC pension report finds that states with relatively higher tax rates often have larger unfunded pension liabilities.xiv

Meanwhile, as stated previously, states are still contractually and constitutionally obligated to pay pension liabilities, so there has been a major divergence between the risk premiums of pension assets and liabilities.

As the Society of Actuaries’ Blue-Ribbon Panel on Public Pension Plan funding recommends, “the rate of return assumption should be based primarily on the current risk-free rate plus explicit risk premium or on other similar forward-looking techniques.”xv This is similar to the blended discount rate recommended by GASB.


The most recent example of the inability for states to alter pension payments comes from a ruling in the Illinois State Supreme Court on May 8, 2015. The state of Illinois passed pension reform in December 2013 state law that stopped automatic, compounded yearly cost-of-living increases, extended retirement ages for current state workers, and limited the amount of salary used to calculate pension benefits. Several public sector unions sued the state of Illinois and on May 8, 2015 the pension reforms were ruled unconstitutional.xvi After rolling back these pension reforms, unfunded liabilities continued to grow and left Illinois one of the worst funded public pension systems in the United States.

Because U.S. Treasury bonds are insured with the full faith and credit of the United States government, the rate of return for these bonds is the best proxy for a risk-free rate. A valuation of liabilities based on a risk-free rate contrasts sharply with the overly optimistic assumptions used by nearly every public sector pension plan. As economist Joshua Rauh notes:

The logic of financial economics is very clear that measuring the value of a pension promise requires using the yields on bonds that match the risk and duration of that promise. Therefore, to reflect the present value cost of actually delivering on a benefit promise requires the use of a default-free yield curve, such as the Treasury yield curve. Financial economists have spoken in near unison on this point. The fact that the stock market, whose performance drives that of most pension plan investments, has earned high historical returns does not justify the use of these historical returns as a discount rate for measuring pension liabilities.xvii

For this reason, it is recommended that states use the lower discount rate. In its annual pension report Unaccountable and Unaffordable, researchers at the American Legislative Exchange Council use a risk- free rate (based on US treasury bond yields) and a fixed discount rate (4.5%) in comparison to discount rates provided in state financial documents. For the 2019 report, a 15-year midpoint, using a hypothetical 15-year U.S. Treasury Bond yield, is used to derive an estimated risk-free discount rate of 2.96%. This is calculated as the average of the 10-year and 20-year bond yields.xviii

The 15-year midpoint comes from the GASB recommendation that a pension plan take no longer than 30 years to pay off its pension liabilities. While state financial documents are not required to report their liabilities projected over a time series (i.e., reporting total liability due per year for the next 75 years), this report must assume the midpoint of state liabilities in order to recalculate state liabilities under different discount rate.xix

This methodology was developed by Bob Williams and Andy Biggs when this report was created by State Budget Solutions, which is now a project of the Center State Fiscal Reform at ALEC. It normalizes the liability values across plans and presents a more prudent valuation of liabilities than many state benefits plans with more rosy assumptions (such as higher discount rates). The inclusion of the fixed discount rate of 4.5%, was added by Thurston Powers in Unaccountable and Unaffordable, 2018.xx

Discount rates used for pension plans can vary even among plans within a state. The use of a risk-free discount rate normalizes discount rates across pension plans, providing the means to assess present value of liabilities across plans. This provides a basis of comparison for liabilities and funding ratios across the 50 states. Other variables provided by state financial documents such as mortality rates, demographics and health care costs were assumed to be correct and not normalized across plans.


This is a more prudent discount rate than many plans offer. The formula for calculating a risk- free present value for a liability requires first finding the future value of the liability. That formula, in which “𝑖” represents a plan’s assumed discount rate, is shown by equation 4xxi:

(4) 𝐹𝑢𝑡𝑢𝑟𝑒 𝑉𝑎𝑙𝑢𝑒 = 𝐴𝑐𝑡𝑢𝑎𝑟𝑖𝑎𝑙 𝐴𝑐𝑐𝑟𝑢𝑒𝑑 𝐿𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑦 × (1 + 𝑖)15

The second step is to discount the future value to arrive at the present value of the more reasonably valued liability. That formula is shown in equation 5 below, where “𝑖” represents either the risk-free or fixed discount ratexxii:


To demonstrate the differences in discount rates, examine the following graphics of the Pennsylvania public pension plans: Municipal Retirement Fund (MRF), Public School Employee Retirement Systems (PSERS), and State Employee Retirement Systems (SERS) for FY 2018 using various discount rates:


The actuarial value of assets is shown on the far left of each chart. Then the chart shows, in appearance from left to right, the present value of liabilities using the “blended discount rate” found in the state’s comprehensive annual financial report (CAFR), followed by the present value of liabilities using the ALEC fixed discount rate of 4.5%, and then finally the ALEC FY 2018 risk-free discount rate of 2.96%. All unfunded liabilities are calculated by subtracting the valued liabilities from the Actuarial Value of Assets.

Using the blended discount rate, the MRF shows a net surplus for FY 2018, however, the ALEC fixed and risk-free discount rates show over $1.1 billion and $2 billion in unfunded pension liabilities, respectively. For the PSERS plan, unfunded liabilities are nearly $94 billion under the fixed discount rate and $132 billion under the risk-free discount rate, significantly larger than the $44.9 billion using the state’s assumed discount rate of 7.25%. For the SERS plan, unfunded liabilities reach $47.4 billion using the fixed discount rate and $66.5 billion under the risk-free discount rate, far greater than the $22.8 billion using the 7.25% discount rate.

The purpose of these charts is to show that assumptions matter, especially the discount rate. Using lower discount rates do not “add” liabilities and using higher discount rates do not “take away” liabilities. The discount rate simply shows the value today of liabilities that are owed in the future. Using a more prudent rate of return, such as the ALEC risk-free rate of return, will provide a clearer picture of pension liabilities because, as previously stated, state governments are legally (and often constitutionally) bound to make those pension payments. The lower discount rate reflects the state government’s inability to default on its pension promises.


Unfunded Liabilities: A nearly $5 Trillion Problem

The ALEC annual report on pensions found that, using a risk-free discount rate, unfunded liabilities totaled $4.9 trillion (a little more than $15,000 per person) for FY 2018.xxiii The following table has been recreated from the ALEC annual report on pensions Unaccountable and Unaffordable with permission from the authors:

As one can see, the unfunded liabilities are not evenly dispersed among the 50 states. The 10 states with the largest unfunded liabilities (the bottom 10 states in the ranking) are Georgia, Massachusetts, Michigan, Florida, New Jersey, Pennsylvania, New York, Ohio, Texas, Illinois, and California. Together these states have a combined sum of $2.97 trillion in unfunded liabilities. This sum makes up 60.27% of all unfunded pension liabilities in the country.

To put these unfunded liabilities in context, it is important to examine unfunded liabilities per capita. That measurement shows the unfunded pension liabilities apportioned out to each state resident. The table below is recreated from the ALEC annual pension report with permission from the authors:

This measurement is concerning because it shows the personal share of liability for every resident in each state. As economist James M. Buchanan noted, government debt is a cost borne by taxpayers in the future.xxiv Each of these unfunded liabilities represents a future tax upon the residents of each state from these unfunded liabilities. States will relatively smaller populations face the largest burdens. Take Illinois for example, where unfunded liabilities are $28,200.06 per capita. That money could pay for tuition, fees, and room and board at Illinois State University for 2020-2021 academic year (just over

$25,000) and have some money left over to pay for books and school supplies.xxv

It is also important to note that these unfunded liabilities make up sizeable portions of their respective state GDPs measurements. The table below is recreated from the ALEC annual pension report with permission from the authors:

Unfunded liabilities as a percentage of state GDP shows a state’s ability to pay its unfunded liabilities. This measurement highlights the severity of a pension crisis in a relatively smaller economy (such as Illinois, Kentucky, or Mississippi). However, large unfunded liabilities can even have a significant impact on relatively larger economies. California, which is a growing economy, but unfortunately unfunded liabilities are also growing at a rapid pace.

As previously mentioned, the funding ratio is an important measure of a defined benefit pension plan’s health. The table below is recreated from the ALEC annual pension report with permission from the authors:

Also, as previously mentioned, states must always strive for 100% funding ratio, however most states fell far below the 100% range. For 2018, the weighted average funding ratio was 34.8%.xxvi This is dangerously low for pension plans. The best funded state, Wisconsin, has consistently had the highest funding ratio among the ALEC rankings since FY 2012. The chart and table below show the 3 highest and 3 lowest funding ratios for FY 2018 over time using the ALEC fixed discount rate of 4.5% compared to the weighted average funding ratio for each fiscal year:


The dashed line in the graph represents the weighted average for each fiscal year. As shown in the chart, the three highest funding ratios for FY 2018 (Wisconsin with 87.92%, South Dakota with 75.37%, and Utah with 70.86%) have consistently performed above the weighted average since FY 2012. Illinois and Connecticut have consistently performed below the weighted average since FY 2012, symptoms of underfunding and poor management. West Virginia only recent performed below the weighted average starting in FY 2017. This was due to the growth of unfunded liabilities outpacing the growth of asset values, controlling for discount rates.

On a positive note, average funding ratios have consistently increased during the period of FY 2012-2018, controlling for discount rates. This is, in part, due to increasing returns in pension assets as well as some state paying down unfunded liabilities and making reforms to make pension systems more sustainable. The next section will discuss how Maine, Michigan, Tennessee, and Wisconsin have made effective changes to their pension systems, keeping them sustainable for the future.

Case Studies for Reform: Maine, Michigan, Tennessee, and Wisconsin

This section will discuss how the states of Maine, Michigan, Tennessee, and Wisconsin have all made workable reforms to their public pension systems in order to keep them sustainable for the future. Among these pension reforms, there are many similarities between the state pension systems.

Wisconsin (as previously mentioned) has the best funded pension system in the country at 70.37%, controlling for difference in discount rates, because it has a variable benefit rate, meaning the disbursement varies over time. State retirees are entitled to a low, guaranteed pension payment paired with a variable payment based on the pension system’s funding ratio.xxvii This means when tax revenue is lower during economic recessions, the fund lowers payments to retirees and allows the fund to recover rather than exhausting the fund or taking on debt to keep making payments.xxviii While the plan has been criticized for diminishing benefits during economic downturns, it has succeeded in providing retirement security with few significant changes to the plan since 1975.xxix

In 2016, Maine pursued a series of reforms to implement variable contribution rates for their state pension system.*** Due to these reforms, in the past two years Maine’s unfunded pension liabilities have decreased by almost $10 billion (about 50%). Normally, employer contribution rates fluctuate to meet the ARC or other contribution standards, whereas employee contributions are a fixed rate set by contract. Under a “risk-sharing” plan, changes in the ARC result in changes in contributions for both employer and employee.

The models share a key aspect: both Maine and Wisconsin have automatic “triggers”, either on contribution rates, benefit rates, or cost of living adjustments. These triggers serve as an objective management tool to ensure pensions are funded. Automatic adjustments based on actuarial science are difficult to argue against, particularly when the potential deviation will underfund the pension system.

In addition, numerous states (e.g. Michigan, Pennsylvania, Wisconsin, and Tennessee) have introduced hybrid pension plans and options for full defined-contribution pensions. In most cases, a hybrid is a relatively small defined-benefit pension plan offered in tandem with a defined-contribution plan. The defined-benefit portion of these hybrids carries all the same risks as traditional pension plans. The risks, however, are mitigated by the smaller size and, often, better contract terms, such as benefit formulas that block spiking (getting large raises or bonuses in the time immediately before retiring in order to get a higher pension payout during retirement) or higher employer contribution rates.

Tennessee currently offers a hybrid pension plan for all state and higher education employees hired on or after July 1, 2014. All state and higher education employees hired before that date have been incorporated in the defined-benefit legacy plan.xxxi The hybrid plan incorporates both a defined- benefit plan and the option to set aside money in a 401(k) plan. Tennessee is consistently one of the states with the best funding ratios and the lowest unfunded liabilities per capita in the ALEC pension reports since 2016. Unfunded liabilities will continue to fall as more retirees participate in hybrid pension plans and the state legacy pension plan liabilities are paid off.

Similarly, Michigan transitioned its Public School Employee Retirement System (MPSERS) to a hybrid pensions for all new hires in 2017. The plan auto-enrolls new hires in a defined-contribution plan, but new teachers have the choice of opting into a hybrid plan with a mix of defined-contribution and defined-benefit plans.The defined-benefit plan splits all costs 50-50 between employers and employees, uses a 10-year amortization schedule and uses a 6% discount rate. In addition, if the hybrid plan’s funding ratio falls below 85% for two consecutive years, the plan is closed to new hires until the funding ratio rises above the 85% threshold for two consecutive years.

In 1996, Michigan was the first state in the nation to close its defined-benefit State Employee Retirement System (MSERS) and enroll new hires in a hybrid plan. However, other state employee plans (such as MPSERS, the State Police Retirement System, State Judges Retirement System, Municipal Employees Retirement System, and the Legislative Retirement System) kept the defined-benefit option open to new hires. Thus, unfunded liabilities continue to accumulate in the other Michigan pension plans.

The case of Michigan demonstrates that a transition to defined-contribution plans does not mean unfunded liabilities will disappear overnight (or even in one fiscal year). Michigan still ranks 42nd in the nation on unfunded liabilities, but the counterfactual would be much worse. If these reforms were not in place, Michigan would resemble its neighbor to the southeast, Ohio, or nearby Illinois. Ohio (48th in the nation) has $290 billion in unfunded liabilities, while Illinois (49th in the nation) has nearly $360 billion in unfunded liabilities. A study conducted by Richard Dreyfuss and the Mackinac Center found that Michigan’s reforms saved taxpayers $167 million in pension liabilities, $2.3 billion to $4.3 billion in unfunded liabilities and improved the political incentives of pension funding. By continuing reforms to transition more pension plans to defined-contribution, Michigan can steadily improve its retirement plans and reduce its unfunded liabilities.

By following the examples of these states, other state pension plans can make improvements to their respective pension systems, keep pensions sustainable for years to come and lower the burden of pension debt on taxpayers.


Further Recommendations: Board Governance Structure and Transitioning to Defined Contribution

This section will explore two further recommendations: Adjusting pension plan board governance structure and making the transition to defined contribution (the structure of most 401(k) plans). The first is a more modest adjust while the second is a complete change to the pension plan.

When discussing the governance structure of pension plan boards, it is important to note that not all pension board structures are created equal. Recent research from Aleksandar Andonov, Yael Hochberg, and Joshua Rauh note that the structure of the pension boards strongly correlates with pension asset investment performance. Andonov, Hochberg, and Rauh find that pension funds whose boards have high fractions of ex officio members (i.e. state treasurers) or members appointed by a state official underperform the most, followed by funds whose boards have a high fraction of members elected by participants.xxxvii The authors noticed that the worst performing governance structures invest more in riskier assets such as real estate and funds of funds, partially explaining the lower performance. Lack of financial experience partially contributes to poor performance by pension boards whose structure is made up of investment but does not explain performance of boards with state officials.xxxviii Political contributions from the finance industry to state officials on pension fund boards are strongly and negatively related to performance, but the authors found it did not fully explain lower performance.

Currently, legislation in the New Mexico state senate would restructure its Public Employee Retirement Association (PERA) Board. The proposed bill (Senate Bill 201), would remove 3 trustees (reducing the total trustees from 12 to 9) with 4 active employees, 2 retired PERA members, New Mexico’s Secretary of Finance and Administration (replacing the Secretary of State and the Treasurer), and add two non-PERA members that have “skill, knowledge, and experience in retirement investment products or retirement plan designs” (i.e. experts in finance and retirement portfolio management).xl Those in favor of Senate Bill 201 cite Andonov, Hochberg, and Rauh’s research on pension board governance structures.

Ultimately, the way to ensure that unfunded pension liabilities are reduced, and large burdens are not placed on taxpayers is to completely transition to defined contribution. The defined contribution pension (a structure like that of the 401(k) plan in the private sector) is a type of plan where employees make contributions to their own personal retirement accounts (and in some cases employers will match retirement contributions up to a certain dollar amount). The funds put into the retirement account are invested in an index fund and retirement savings accumulate value through investment returns and contributions. Upon retirement, it is up to the retiree to budget what he or she has saved and invested while working.xlii While a defined contribution plan does not promise the annual payments of a defined benefit plan, it offers employees greater flexibility. The employee does not have to wait to be vested in a defined contribution retirement plan and follows the employee if he or she chooses to change careers. That flexibility can allow public sector workers to leave his or her public sector job for another job (whether in the public or private sector) and not worry about losing retirement savings. The defined contribution will follow him or her wherever their career takes them all the while lowering the cost of retirement payments for taxpayers.



Williams, Jonathan; Siconolfi, Kati; Lafferty, Ted; and Young, Elliot. “Keeping the Promise: Getting Politics out of Pensions.” American Legislative Exchange Council. 14 Dec. 2016. Retrieved from:

Weinberg, Sheila and Norcross, Eileen. “GASB 67 and GASB 68: What the New Accounting Standards Mean for Public Pension Reporting” Mercatus Policy Briefs. 15 June 2017. Retrieved from: mean-public-pension

“Notes to Required Supplementary Information” Illinois Comprehensive Annual Financial Reports (Numerous Pension Plans). 2018. Retrieved from:, Jonathan; Savidge, Thomas; Schalk, Lee; and Williams, Bob. Unaccountable and Unaffordable: Unfunded Pension Liabilities in the 50 States. American Legislative Exchange Council. 2020. (Forthcoming).

Williams, et. al. Unaccountable and Unaffordable (Forthcoming).

Weinberg and Norcross, “GASB 67 and GASB 68.”

“The 80% Pension Funding Standard Myth.” American Academy of Actuaries Issue Brief. July 2012. Retrieved from:

Weinberg and Norcross, “GASB 67 and GASB 68.”

“Wyoming Air Guard Firefighters’ Retirement System Actuarial Valuation Report for the Year Beginning January 1, 2019.” Wyoming Retirement System. 18 April 2019. Retrieved from:

“State Actuarial Valuation As of June 30, 2018” CalPERS. 2019 October. Retrieved from:

Williams, et al. Unaccountable and Unaffordable 2019 (Forthcoming).

Williams, et al. “Keeping the Promise.”

Williams, et al. Unaccountable and Unaffordable 2019 (Forthcoming).

Report of the Blue-Ribbon Panel on Public Pension Plan Funding. Society of Actuaries. 2014. Retrieved from:

In re Pension Reform Litigation, 118585 IL. 2015. Retrieved from:

Rauh, Jonathan D. Testimony Before the Joint Select Committee on Solvency of Multiemployer Pension Plans. United States Senate & United States House of Representatives. July 25th, 2018.

Williams, et al. Unaccountable and Unaffordable 2019 (Forthcoming).

xx Powers, Thurston; Smith, Christine; Williams, Bob; and Williams, Jonathan. Unaccountable and Unaffordable 2018. American Legislative Exchange Council. 2018. Retrieved from unaffordable-2018/

Williams, et al. Unaccountable and Unaffordable 2019 (Forthcoming).

Buchanan, James M. The Collected Works of James M. Buchanan Volume 2: Public Principles of Public Debt: A Defense and Restatement. Indianapolis: Liberty Fund (1999). p. 22.

“Cost of Attendance Tables: New, undergraduate students for Fall 2020, Spring 2021, or Summer 2021. Financial Aid Office of Illinois State University. 2020 March. Retrieved from:

Williams, et al. Unaccountable and Unaffordable 2019 (Forthcoming).

Seyfert, Kerri. “The Wisconsin Retirement System Is Fully Funded.” Reason Foundation. Jan. 2020.


Stein, Jason. “Wisconsin’s fully funded pension system is one of a kind.” Milwaukee Journal Sentinel. 26 Setp. 2016. Retrieved from: one-of-a-kind.html

Braun, Martin Z. “Public Pensions Adopt Cost Sharing Mechanisms to Stem Volatility.” Bloomberg. July 17, 2018. Retrieved from: mechanisms-to-stem-volatility

“GASB and Actuarial Information” Tennessee Department of Treasury. Updated 2019. Retrieved from:

Gilroy, Leonard, Randazzo, Anthony, and Takash, Daniel. “Michigan Adopts Most Innovative Teacher Pension Reform in the Nation.” Reason Foundation. 16 June 2017. Retrieved from:

Randazzo, Anthony. “Pension Reform Case Study: Michigan 2016 Update.” Reason Foundation. Aug. 2016.

Retrieved from:


Dreyfuss, Richard C. “Policy Brief: Estimated Savings from Michigan’s 1997 State Employees Pension Plan Reform.” Mackinac Center for Public Policy. 23 June 2011. Retrieved from:

Andonov, Aleksandar; Hochberg, Yael V; and Rauh, Joshua D. “Pension Fund Board Composition and Investment Performance: Evidence from Private Equity.” Hoover Institution Economics Working Papers. March 2016. Retrieved from:

Niraula, Anil; Gassenberger, Steven; and Gilroy, Leonard. “Commentary: Proposed New Mexico PERA Board Restructuring Would Improve Expertise, Balance Representation Long-Term.” Reason Foundation. 5 February 2020. Retrieved from: balance-representation/

Williams, et al. Unaccountable and Unaffordable 2019 (Forthcoming).


Pension Liability Stress

Leave a Reply