Breaking Down the Meltdown

From our January-February 2014 edition of Fiscal Focus.  The latest state and local pension valuation reports now show a collective deficit of $17 billion.  We break down how we got to this point by reviewing valuation report histories and assigning numbers to the various contributing factors.

Once upon a time the phrase “public sector pension benefits” was guaranteed to trigger nothing more than a yawn. Hot off an incredible decade of investment returns, the majority of the big state and local pension plans (see Table 1 for list) were “overfunded” (assets exceeding liabilities) despite having handed out annual base benefit increases to retirees for several years that often approached 10% or more. According to the July 1, 2001 valuation reports, the largest state and local plans collectively had $187.2 million in assets in excess of liabilities.

Table 1

Public Pension Plans Included in Analysis

  Statewide General Employee Plans Notes
  MSRS General  
  PERA General
  Statewide Specialty Plans Notes
  MSRS State Patrol  
  MSRS Correctional  
  MSRS Judges  
  PERA Police & Fire  
  PERA Correctional  
  PERA MERF Division Beginning July 1, 2010
  Local Plans Notes
  Minneapolis Employees (MERF) Through June 30, 2010
  Minneapolis Teachers (MTRFA) Until July 1, 2006 merger with TRA
  Duluth Teachers (DTRFA)  
  Saint Paul Teachers (SPTRFA)

Twelve years later, based on the recently released valuation reports, these plans collectively now have a tad over $17.25 billion in unfunded liabilities. Two of the major local plans no longer exist having been absorbed into “healthier” state plans that are struggling with their own funded status. Two other major local plans are in significant financial distress and in a sort of limbo as the legislature decides what to do with them.

To better understand how and why the unfunded liabilities grew so far so fast, we reviewed the past twelve years of actuarial valuation reports and assigned the growth in unfunded liabilities to the various contributing factors. Among other things, actuarial valuations measure the extent to which various factors generate either actuarial losses (adding additional unfunded liabilities) or actuarial gains (eliminating unfunded liabilities) when actual experiences deviate from the assumptions under which the plans operate.1

The chart below shows how selected factors contributed to the $17.25 billion increase in unfunded liabilities over the period we are considering. Anything above zero added to the unfunded liability; anything below zero reduced the unfunded liability. Following is brief description for each bar on the chart, along with a few comments and takeaways.

Figure 1

Cumulative Change in Consumer Price Index and Implicit Price Deflator for State and Local Governments: 1973 to 2012

Contribution deficiency: In Minnesota, pension contributions come from employees and employers, occasionally supplemented by state aids. The contribution deficiency captures the extent to which these contributions failed to cover the cost of three items: 1) new benefits accrued during the year or “normal cost”; 2) administrative costs associated with operating the plan; and 3) the cost of amortizing any unfunded liabilities. Over the twelve-year period in question the collective contributions made by public employees and government did not cover these costs – creating $4.53 billion in additional unfunded liabilities.

Contribution deficiencies have impacted PERA General, which covers most local government employees, especially hard. It is the only plan to have a contribution deficiency in each of the 12 years under consideration. Importantly, this in spite of an increase in employer contributions from 5.18% to 7.25% of salary (a 40% higher rate) and employee contributions increasing from 4.75% to 6.25% of salary (a 32% higher rate) since July 1, 2001. All three statewide general employee plans – which cover the bulk of the state’s public employees – have seen their employee and employer contribution rates increase by at least 20%.

Salary increases: The normal cost for an additional year of benefits is computed using certain assumptions, one set of which relates to salary growth. If salaries grow faster than anticipated, more money is needed to fund benefits, and there is an actuarial loss as actual costs come in above projections. Conversely, if salaries grow more slowly than anticipated, less money is needed to fund benefits and there is an actuarial gain. Influenced by two recessions during this period – one not so great, the other “Great” – government salary growth overall was consistently lower than expected, resulting in an actuarial gain for these retirement plans of $3.92 billion on a combined basis, offsetting some of the overall growth in unfunded liabilities.

Investment Income: When are near-all time record highs in investment markets not enough? The answer: when you are dependent on more than what those markets have given you. Investment income captures (in this instance) the collective impact of failing to meet the state’s expected investment returns of 8.5%2 annually. It is clearly the largest contributor to the actuarial losses, and the main driver in changes to these retirement plans’ financial health. Over this twelve-year period, sustained failures to hit the 8.5% return target created a staggering $22.69 billion in additional unfunded liabilities. To provide some sense of perspective on both the expectations and the “miss,” an 8.5% compounded return on the Dow Jones Industrial Average beginning in July 2001 would have the Dow at 28,120 today.

Changes in plan provisions: This measures how changes in plan benefits affected unfunded liabilities. In other words, this mostly captures the impact of the oft-heralded “sustainability fixes” which were enacted a few years ago – much of which dealt with reducing or eliminating annual cost of living adjustments. All told, changes in plan provisions over the years put a dent in the problem by reducing unfunded liabilities by $6.38 billion.

Changes in actuarial assumptions and methods: This factor measures the extent to which changes in actuarial assumptions or methodologies employed in the valuations affect the level of unfunded liabilities. The drawn out saga and process which eventually eliminated the “Post Fund” – a fund which segregated assets for retirees drawing pensions from the statewide funds – triggered actuarial changes that resulted in a $1.08 billion reduction in unfunded liabilities.

Other factors: The black box of pension valuations is chock full of other assumptions about what the future holds. ”Other factors” measures the extent to which reality mirrored these economic and demographic expectations. It includes gains and losses relative to expectations about retirement ages, how long employees work, when beneficiaries die, and many other items. Over time, the deviations from these assumptions have generated $1.42 billion in unfunded liabilities.

What to make of all this? For starters, it brings into sharp relief how the calculation of unfunded liabilities and contribution deficiencies – which determine how the public perceives pension plans – is so dependent on decisions made by policymakers. Given the special influence assumed investment returns and chosen timeperiods for paying off unfunded liabilities have on reported health, governments exercise considerable control over the presentation of pension reality they want to give to taxpayers.

The link between the investment income and the required contributions – higher expected returns lower the contributions needed – puts Minnesota’s chronic failure to make the necessary contributions in an interesting light. There’s a powerful incentive to lower current pension costs by assuming the money will work harder which Minnesota pursues with a vengeance using an expected return of 8.5%. Yet we still failed to make the necessary contributions – $4.5 billion short across all governments over 12 years – even though the most aggressive return expectations in the nation helped drive contribution requirements about as low as they could possibly go.

Finally, the difficulties of investing our way out of this are pretty clear. An 8.5% return is essentially treading water – making up $17 billion on top of that is a tall order for the state’s investment managers. As a rough illustration, think back to the Dow Jones Industrial Average numbers we quoted before. As a point of reference, an 8.5% compounded return beginning in July 2001 would have the Dow at nearly 81,000 by the end of June of 2026 – another 13 years down the road. To get there from where we’re at, the pension plans need to realize returns of nearly 14% for the next 13 years. Meanwhile any investment shortfall – even if it’s a positive return that’s less than 8.5% – creates new actuarial losses making the journey back to full funding that much more difficult.

  • 1 The data is incomplete in one respect: 23 of the 136 actuarial valuations we studied did not isolate the actuarial gains or losses generated by salary increases.  We requested the information from the pension funds but were told in most cases it did not exist, although we did receive additional data for two of the valuations.  The missing salary-related actuarial gains and losses is included in "other factors".
  • 2 Technically the assumed return is 8.0% for the next four years then resuming its normal 8.5% in perpetuity thereafter.  The temporary 5 year decline was a negotiated measure to demonstrate some sensitivity and action on this issue.