ICYMI | State and Local Government Pensions at the Crossroads

Heightened public awareness.

Beyond the significant attention that GASB Standards 67 and 68 bring to measurement and reporting, the impact of the changes has been an impetus for public debate, plan review, policy conversation, and media coverage, often with close attention to underfunding. This issue is not new; however, concerns may be greater given that the typical funding percentage dropped approximately 10% in 2008–2010 due to investment losses during the recession. As of this writing, there has yet to be a full recovery.

Exhibit 2 depicts the percentage of aggregate state government pension liability. In 2014, the total liability was $3.7 trillion, of which $900 billion was unfunded. Recent conversations and high-profile concerns in certain cities and states can in part be attributed to improved disclosures. Increased public interest is directed particularly at the amount of the liability and the discount rate.

Blended rate increases unfunded liability.

The blended discount rate reflects the long-term expected rate of return on plan assets achievable following the plan’s investment strategy and forecasted employer, non-employer, and employee contributions sufficient to cover projected benefits for currently covered employees; to the extent these conditions are not met, this is blended with the yield rate for 20-year, tax-exempt general obligation municipal bonds with an average rating of AA/A or higher.

When applied to the entire government pension sector using the Pew Charitable Trust Pension Database (The State Pension Funding Gap 2014: New Accounting Rules Help Provide a Clearer Picture, Pew Charitable Trusts, August 2016, http://bit.ly/2n0SLLl), the unfunded liability increases dramatically under GASB’s blended rate scenario. In addition, changes in municipal bond rates, as well as a range of issues regarding future funding, reflect future risks and influence the blended discount rate. Exhibit 3 evaluates the impact of using alternative discount rates. The average discount rate in 2014 was 7.3%, which translates to a total unfunded liability of $930 billion. That rises to $1.57 trillion when the discount rate is dropped to 4%, decreasing the funded percentage from 75% to 64%. If, as originally proposed, a less risky rate such as 4% were applied to the entire liability, the unfunded liability would increase from $900 billion to $3.41 trillion and the funded percentage would drop from 75% to 45%. How the liability changes based upon use of alternate discount rates highlights the vulnerability of pension funds to economic conditions.

Expanded auditor testing.

Auditors are now testing new data regarding not only the net pension liability, but also deferred inflows, deferred outflows, and pension expense. These assessments are challenging matters regarding audits of local governments under a state or multi-employer plan. The data that local government auditors start with is in most cases derived from state plans, particularly regarding retirees and inactive employees, active employees, and actuarial data. A review and reconciliation of the local government retiree data with that of the state plan should be prepared each year.

Ratings agencies adjust.

An initial concern was that there would be massive bond ratings downgrades and the cost of debt for state and local governments would soar if GASB’s measures were too severe. Such ratings intuitively should dip with greater liabilities and new disclosures that signal greater under-funding; however, a review of state bond ratings over 10 years shows little movement (Pamela Prah and Stephen Carr, “Infographic—S&P State Credit Ratings, 2001–2014,” Stateline, June 2014, http://bit.ly/2ndNL6E). Moody’s made four adjustments that better aligned its analysis to GASB Statements 67 and 68:

• Multiple-employer cost-sharing plan liabilities were allocated to specific government employers based on proportionate shares of total plan contributions;
• Accrued actuarial liabilities were adjusted based on a high-grade, long-term taxable bond index discount rate as of the date of valuation;
• Asset smoothing was replaced with reported market or fair value as of the actuarial reporting date; and
• Net pension liability was amortized over 20 years using a level-dollar method to create a measure of annual burden related to the net pension liability. (Adjustments to U.S. State and Local Government Reported Pension Data, Moody’s Investor Services, April 2013, http://bit.ly/2m1q81j.)

Standard & Poor’s (S&P) response was to accept the movement of the liability to local governments and analyze accordingly. S&P continued to focus on commitment to funding, investment performance, trend analysis, and overall sustainability based on state management (John Sugden, “Standard & Poor’s Approach to Pension Liabilities in Light of GASB 67 & 68,” Ratings Direct, July 2013, http://bit.ly/2m1sKwo), stating, “We do not anticipate significant revisions to state rating solely on the changes to GASB reporting.” Similarly, the new GASB rules have had little bearing on Moody’s credit ratings because they already factor the unfunded liability in their calculations. Moody’s Thomas Aaron said, “While GASB 67 and 68 impose many new rules related to pension accounting disclosure, our approach to evaluating credit risk stemming from public pensions remains fundamentally unchanged” (“Modest Credit Impact for GASB Pension Changes, but Con tribution Weaknesses Now High lighted,” Moody’s Investor Services, Mar. 16, 2015, http://bit.ly/2ntLOAb). But Moody’s also observed that weaknesses are now better highlighted.

The analysts seem to believe that, although the extent of pension underfunding was news to the public, the rating agencies had already been doing their own calculations and adjusting for underfunding. Nevertheless, S&P noted that “the changes to pension liabilities resulting from the new GASB standards, such as the use of the blended rate, are more likely to affect governments for which we have already factored their weak pension funding status into our ratings” (Sugden 2013).

Liability shifting to local government.

Local government units that are part of a state or other multiemployer plan must now report directly on their financial statements a portion of the liability for retired employees collecting from the state pension fund and current employees who have earned pension benefits directly on their financial statements. Despite the bond market’s adjustment for this change, the public and many government officials seemed unaware; even mid-sized governments are now seeing liabilities in the hundreds of millions. Many if not all state and local government entities have had in-depth meetings between management and various government legislative and advisory bodies on the issue. Exhibit 4 highlights the redistribution of the liability from states to local governments (Munell and Jean-Pierre Aubry, “Will Pension and OPEBS Break State and Local Budgets?” State and Local Pension Plans, October 2016, http://bit.ly/2ntoORM). Note that states have numerous mechanisms in place, described below, that could be exercised before any liability payment would need to be assumed by local agencies.

Shifting the cost.

Cost shifting from pension funds to individuals and employers is a technique revealed by data compiled by the National Association of State Retirement Administrators (NASRA); their research data, analysis, and briefs on a variety of government pension issues can be found at http://bit.ly/2n34d9b. Specifically, NASRA data shows that employees are contributing more to funds at higher rates, COLAs are decreasing, and employers are paying more.

Exhibit 5 provides two maps of states that 1) established COLA decreases and 2) changed employee contributions. Thirty-eight states have increased required employee contributions since 2009. Several states, such as Pennsylvania, Arizona, Iowa, Kansas, and Nevada, have implemented variable rates that automatically increase based on actuarial estimates.

COLA adjustments have long been offered to retirees to help their income keep up with inflation. Over the course of a retiree’s life, even a small gap between inflation and COLA will reduce benefits. The standard used to be a guaranteed fixed amount that would be adjusted up for higher inflation; today, high-end COLA plans adjust for inflation with a fixed cap. Since 2009, fifteen states have made adjustments impacting current retirees, eight have reduced benefits for all future retirees who are current employees, and seven have adjusted for all new employees. COLA reductions come in many forms, including direct percentage cuts, eliminating the guaranteed increase amount, delaying adjustments for a period of years or until a retiree reaches a certain age, and limiting COLA to a smaller portion of a pension.

Increasing general public funding.

Chicago chose a tax hike (and Dallas is considering one), while California and other states are raising additional revenue by requiring non-government employees to join the system and contribute. This method may help appease current members and some of the unions but risks provoking the general public. Wolf Richter worries that “regular families who’re just sitting ducks are going to get squeezed dry, in order to slow the momentum of the public-employee pension crisis” (“U.S. Pension Crisis: This Is How Families Get Squeezed to Bail Out Pension Funds in Chicago,” Wolf Street, Nov. 17, 2016, http://bit.ly/2m57l59). Another criticism concerning California’s approach is found in the rhetoric of the Federalist: “These same irresponsible politicians have figured out a way to bail out their government works pension systems: forcibly enlist the general public into underwater systems, taking their money” (DeVore 2016). Finally, the California Supreme Court has agreed to hear the issue of whether pension benefits can be cut.

Increasing state funding.

Some states have opted to pick up the cost of reducing the unfunded liability themselves, with changes in funding policies in Tennessee and West Virginia; Alabama, however, chose not to follow this approach (“The State Pensions Funding Gap: Challenges Persist, Pew Charitable Trusts,” Jul. 14, 2015, http://bit.ly/2ndTxVS). In South Carolina, the legislature has considered a package of solutions, including spending more taxpayer money, but also taking more from government workers and suggesting defined contribution plans.

Seeking higher returns.

Achieving higher investment returns, net of inflation, would of course be helpful; however, optimistically anticipating higher earnings and failing to achieve them risks repeating the problem. In Morningstar‘s user return expectation survey, most respondents did not foresee the high returns that would be necessary (Christine Benz, “What Market Experts Are Saying about Future Returns,” Jan. 14, 2016, http://bit.ly/2nw44ck).

Reducing benefits.

Lower pension benefits are another option. They could come in the form of COLA reductions, formula changes as in Ohio, or as direct benefit reductions as in Detroit. Detroit’s court ruling may provide precedent as to the limits of such an approach, but it does not provide a clear solution. Furthermore, reductions, while substantial to individual retirees, may have only a modest impact on fund deficits.

Expanding the federal government role.

In the past, federal legislators sought to aid private sector employers such as airline or steel industries. The Pension Benefit Guarantee Corporation (PBGC), however, lacks the mandate to go beyond the private sector. Among the reasons the federal government remains a bystander, in addition to complex state’s rights issues, is the perception and concern over “long-standing poor governance and planning, profligacy, continual log rolling, and as reported in the media, sometimes significant conflicts of interest with investment firms, legislators, and unions” (Mark Warshawsky and Ross A. Marchand, “The Extent and Nature of State and Local Government Pension Problems and a Solution,” George Mason University, January 2016, http://bit.ly/2mT4vz1). Ed Bachrach of the Wall Street Journal has suggested that Congress should pass laws allowing states and local governments to reduce benefits to clear up the uncertainty once and for all (“How to Save Public Pensions, No Federal Bailout Needed,” Jul. 17, 2016, http://on.wsj.com/2n34RUn). Given the current political landscape, however, the involvement of the federal government seems unlikely.

Moving pensioner populations.

Transferring existing and future employees into defined contribution plans, while fiscally appealing, is not politically inviting. Some versions of this approach could also involve a payout of a present value of benefit payments, which may go the furthest to solve the funding problem but lacks a viable transition mechanism. Which present value rate to use, how to effect the transfer, and whether it applies to all employees are just some of the issues. In addition, the psychological sense of income security loss by retirees must be addressed. A voluntary buyout that offers attractive terms is one proposal; in Dallas, pension plan members have withdrawn over $200 million from their fund, concluding it is better to invest on their own than risk bankruptcy. The likelihood of this approach bringing a resolution on a general scale seems limited, however, when examining the finite mathematics involving large-scale withdrawals.

Phasing out defined benefit plans.

Requiring new employees to enroll in defined contribution plans may become a prominent option. In August 2016, the Tennessee Valley Authority Retirement System (TVARS) approved a plan for all employees with 10 years of service or less to move from a defined benefit to a defined contribution plan (Board minutes, Aug. 8, 2016, http://bit.ly/2nrBc8e). Michigan has adopted this plan for all new hires for general state and local workers, but not for its school employees. The Alabama Policy Institute recommended immediate transition of new hires to defined contribution plans and also encouraged existing employees to voluntarily do the same (Kilroy 2016). This approach may ease current fiscal stress but, it is an incomplete solution.

Exploring hybrid models.

Hybrid models may be fertile ground for compromise. They offer a smaller guaranteed portion with a matching defined contribution program element. For example, under a defined benefit program, a retiree making $100,000 a year who works 30 years of service with a multiplier of 2.5% for each year of service would have an annual pension of $75,000. A hybrid model might cut the multiplier to 1%, for example, reducing the annual pension to $30,000. A 401(k)-type matching program would help make up the differential. Georgia, Utah, Tennessee, and Virginia already have these types of plans in place, but only for new employees. In Rhode Island, after several years of court battles, the government settled with labor forces, adopting a hybrid model for both new employees and current employees. This approach may, in due time, be capable of resolving under-funding without raising taxes.

None of these proposals will fit every situation, and none are easy or simple. Some have stronger portability features than others, which may be important in a dynamic employment market. Solutions will vary depending on the culture of the state and the ability for all parties to share the sacrifice. It is more difficult in some states to alter plans due to labor, policy making, and legal constraints. In addition, political elements—which party holds power, Supreme Court appointments, and strength of labor representation—will weigh heavily on decisions. Because of this, it seems unlikely that lawmakers will buy out defined benefit plans as some foresee. Rather, solutions may include offering new employees defined contribution plans or adopting hybrid plans. combined with moderate versions of the solutions outlined above, including larger employee contributions.