Long-Term Liability Challenges for South Carolina Taxpayers
South Carolina taxpayers currently face serious long-term financial challenges based upon a review of the July 1, 2008actuarial report of the largest statewide pension plan – the South Carolina Retirement System (SCRS). This plan covers 192,820 active members in addition to 259,560 non-active members.
South Carolina taxpayers are facing compound problems regarding the state’s ability to effectively manage both defined benefit pension and retiree medical liabilities. While payments to current retirees are not in jeopardy, the ongoing costs to both members and taxpayers will be determined by future asset growth and favorable health care costs — relying on either trend presents significant risks.
SCRS Financial Position
An indicator of this pension problem is reflected in the measure of the SCRS assets  to actuarial liabilities  which was 69.7% as of July 1, 2007. In dollar terms this equates to a deficit of $10.2 billion. The standard for plans should target funding at the 100% level. Such a standard is often disputed by those who believe that an arbitrary 80% ratio is appropriate with the balance to be funded at some future point by some future taxpayers, together with other deferred costs including retiree medical obligations. Some actually maintain that running perpetual pension deficits is, in fact, a cost savings strategy since it precludes politically minded policymakers from considering further benefit improvements.
Notwithstanding all these considerations, the formula maximum governing annual SCRS retiree COLA benefits was amended in 2008 by increasing this limit from 1% to 2% per year. The July 1, 2008actuarial valuation indicates this action alone created over $2.6 billion in new plan liabilities. Of note, this formula will not result in a pension COLA in 2009 since there was not a sufficient rise in the CPI. The methodology used to determine the actual CPI remained unchanged from that used prior years. However, with significant budgetary deficits at the state and federal levels, it is very likely inflation will be a significant issue for the foreseeable future, thereby increasing the likelihood of future COLA payments.
Effective July 1, 2008, The State Budget and Control Board (which is the governing entity for the pension system) also approved certain revised actuarial assumptions and methods – in generic terms “ financial engineering” to help offset the increased cost of this newly increased annual retiree COLA. The net result is more risk is now placed on taxpayers given this benefit improvement coupled with the assumption of higher expected investment returns to “pay” for these increased benefits. The annual investment return assumption was increased from 7.25% to 8.00%.
Finally, to mitigate the significant declines in asset values measured in the year ending July 1, 2008, the revised approach is to determine asset values effectively using a ten-year average. The “Public Fund Survey” released in November, 2008 reports that 87% of similar funds from other retirement systems have a smoothing period of 5 years or less while only 3% of other public funds use a ten year or longer average. It appears this move in South Carolina was done to defer recognition of asset losses and maintain the appearance of level and uneventful taxpayer contributions in the current economic downturn.
As a result of all these changes, the funded ratio as of July 1, 2008on the revised methodology was 69.3% resulting in an increase in the unfunded liability to $11 billion. This figure does not reflect any asset losses since that point as those will be reflected (and deferred) in subsequent reports. Obviously, any market gains will be treated in a similar fashion.
It is noteworthy that in Missouri, state law precludes any benefit improvements in any plan funded at or above the 80% level, if such changes would result in a funded ratio of 75% or less. It is a worthwhile question to understand what the SCRS funded ratio would be on July 1, 2008, if the July 1, 2007actuarial methods and assumption had remained unchanged. In addition, funded ratios based upon the market value of assets would provide interested parties with additional information on the financial position of the fund.
Actuarial assumptions often involve projections of over thirty years. The reality is no one even knows what future returns will be. Therefore, debating whether the correct annual investment return will be closer to 7.25% or 8.00% is at best an inconclusive and theoretical financial exercise. Actuarial assumptions should also be viewed in the aggregate, as these collective assumptions determine the pattern of how plans are to be funded. However, such assumptions do not define the true cost of the plan given this is only measured through actual experience. Many observers do not appreciate this distinction, as they confuse pension contributions with the actual pension costs. Pension funding should be viewed as a pattern of annual deposits towards current and future expenditures made by the plan.
Nonetheless, this change in the investment return assumption should be viewed as placing higher, unnecessary and potentially unaffordable risks on the taxpayer. After all, what was the downside risk to taxpayers in leaving the assumed rate unchanged at 7.25%? Moreover, any empirical analyses justifying a higher rate are likely based upon assumptions quantified through a financial modeling process. This modeling is typically based upon historical norms. The appropriateness of any such modeling can be debated. Such an analysis usually yields a “reasonable range” rather than a finite number. Therefore, it is likely both 7.25% and 8% were within this range.
Among large public pension plans, the observed range of investment returns is between 7% and 8.5% with an average of 8% based upon data from the most recent “Public Fund Survey”. However, these studies often involve a reporting lag of one to two years which therefore does not properly reflect any recent actions by plans regarding the long-term implications of the current economic downturn. The observed trend is to decrease the assumed rate of return which will likely eventually show a decline from the 8% survey average. While the assumption change may not necessarily be “wrong”, the higher assumed rate creates a risk that is not only unnecessary, but also counterintuitive.
Given all this, most financial forecasts are appearing less optimistic in the future. This author is not aware of any plan except SCRS which has actually recently raised its investment rate-of-return assumptions.
Perhaps the answer, at least in part, rests with the increased liabilities associated with the recently approved COLA improvement and the implicit desire of policymakers not to materially increase funding into an already significantly underfunded plan. This action raises important questions regarding the pension funding policies for both the short and long-term and the impact of such a strategy on future taxpayers. The increase of the asset averaging period from 5 years to 10 years also compounds the problem. Such a strategy is counter to the practices of the vast majority of other retirement systems – again, only 3 percent of plans rely on a period of 10 years or longer. There seems to be no rationale for South Carolina’s decision to change the average than to further defer funding increases associated with observed and anticipated poor investment returns. Any such amounts not currently contributed are effectively a loan from the plan which must be repaid at 8% interest per annum. Given that such funding increases are considered unaffordable now, there is little basis on which to assume they will they be affordable in the future, particularly when the interest cost on these unfunded liabilities is considered.
Furthermore, if there were reason to assume that future investment returns would increase, that trend would be a sound basis for contributing significantly more into the plan to make progress on this ever-increasing unfunded liability.
These collective changes raise legitimate concerns regarding the long-term funding policies of the SC Retirement System and the liabilities that will inevitably be assumed by taxpayers.
Teacher and Employee Retention Incentive (TERI)
A provision within SCRS is the Teacher and Employee Retention Incentive (TERI) Program. A summary is this provision is presented below based upon the July 1, 2008 actuarial valuation report.
Age 60 with 5 years of service, or age 55 with 25 years of service, or 28 years of service
Teachers and Employees Retention Incentive (TERI)
Upon meeting retirement eligibility, a member can elect to enter the TERI for a maximum of five (5) years, after which employment will cease. The retirement benefits will be accumulated in TERI accounts and will be paid to the members upon the earlier of actual retirement or the end of participation period. The amount credited to the TERI account is based upon the calculation and form of benefit selected by the member at TERI entry. COLAs are credited to the TERI account.
No interest is credited to the TERI account. Employee contributions (for those entering TERI on and after 7/1/2005) and employer contributions continue during TERI participation.
The TERI program increases pension liabilities versus a similar plan where this provision did not exist. It would be appropriate for actuaries to more carefully analyze the program and quantify the net difference in the liability created by TERI. It is important to recognize that the true cost is always based upon experience. In addition, such an analysis should include the cost of retiree medical benefits.
Some analyses and general opinions which conclude that the TERI creates no additional pension liabilities are flawed in that they fail to properly account for the time value of money.
The fact that TERI exists at early ages (conceivably prior to age 50) makes the early retirement benefits very generous. The retirement system is, in effect, competing against the active workforce for the same employee. TERI allows employees to effectively retire while remaining in the workforce for five more years.
It is atypical to create such a strong and presumed costly early retirement incentives, but advocates of defined benefit pension plans argue it enables state government to attract and retain valued employees for the long-term. Actually, TERI creates a strong incentive for an employee to leave the workforce in mid-career, which creates conflicting goals within the plan. Logically, many individuals subsequently join another employer for the duration of their career until a normal retirement date.
Advocates of such plans justify higher salaries paid to long-service employees as recognition of acquired experience and expertise. Ironically, proponents justify generous early retirement incentives as creating “savings” to the employer. They argue that retiring employees are replaced with new employees at lower pay rates. This line of reasoning suggests all employees are interchangeable regardless of experience. If that is the case, it becomes difficult to defend pay practices, and if it is not the case, then doubt can be cast about the perceived savings.
Based upon newspaper reports in The State (Columbia, SC; 1/10/2009), “Sanford proposes ‘toughest’ cuts for S.C.”
“State agencies would be encouraged not to rehire workers who had finished their five-year contracts in the Teacher and Retirement Initiative, or TERI”
The term “encouraged” suggests that not all TERI members retire following the duration of their contractual period. Such a practice, as reported, turns TERI into a mid-career supplement and would appear to create many HR, legal and actuarial issues. The actuarial issue is significant as the valuation report provides no guidance that any liability provision is made for these “continuing” members.
All this suggests that TERI is a generous answer in search of an undefined problem. In an era where pension and retiree medical liabilities are already problematic for taxpayers, it would seem hardly appropriate to be unnecessarily creating such liabilities.
Moreover, the logic of continuing this plan is confounding in an era of relatively high unemployment where working to Social Security eligibility (age 62) should be encouraged. Pension and retiree medical costs represent a most significant and growing taxpayer expenditure which should be debated by policymakers and justified to taxpayers.
Retiree Medical Liabilities
Equally important should be the issue of managing liabilities for retiree medical benefit plans. Public entities have not been required until recently to account for these future liabilities in a manner generally similar to pensions. The Government Accounting Standard Board Statement 43 and 45 (GASB 45) now requires such accounting standards be adopted. However, these standards do not require public entities to actually prefund these obligations. The author did not review any GASB 45 actuarial reports in connection with this study. The state is reporting an unfunded GASB 45 liability of $8.6 billion which needs to be considered as part of any benefit package afforded retirees. Responsibly managing this liability should also be an important priority to policymakers.
Therefore, the merits and affordability of an enhanced pension COLA can also be debated especially in the context of existing GASB 45 liabilities. Moreover, in a review of 9 major SC private-sector employers ranging in size from 4,500 to 17,700 employees, only 5 sponsored a defined benefit plan and none of the plans included COLAs. In the private-sector such provisions are considered unaffordable, and thus they are even less justifiable in the public sector.
Of note, the majority of public defined benefit pension plans requires significant employee contributions and maintains formula driven or ad-hoc COLAs. These plans generally provide a higher level of benefits compared to the non-contributory private-sector defined benefit plans. Equally true, the net result is most public pension plans remain poorly funded often with significant deferred costs that will likely prove unaffordable to taxpayers. To rationalize this, some cite the arbitrary goal of being 80% funded – that is targeting a 20% current deficit as acceptable.
SCRS Funding Policies
The economic downturn reflected in the asset values has only exacerbated the funding challenges resulting in new strategies to further defer proper funding as exemplified by the SCRS. The SCRS funded status was not good even before the financial downturn. These recent changes have further weakened the plan’s financial solvency. The defense is often comprised of benchmarking to even worse-funded plans in an attempt to mollify taxpayers. Such benchmarking is one of the systematic problems plaguing the public pension system.
Most unfunded liabilities within SCRS are funded over a period up to thirty years which is another commonly-used liability deferral technique. Sound actuarial funding practices are intended to have the aggregate benefits to the group fully-funded or “paid–up” at retirement. To illustrate the problems facing SCRS, the average age of the active membership is 45 (44.67 to be exact). If one assumes individuals retire at an average age of 60, this suggests an amortization period for the unfunded liability of no more than 15 years. Moreover, the SCRS funding policies are expressed as a constant percentage of pay meaning that the liability is paid off as salaries increase. This backloads and serves to further defer costs. In addition, the pension contribution which is expressed as a percentage of payroll cost is effectively assumed by new members given current members, on the average, will retire well before the current 29 year period is completed. Finally, each year resets the calculations, so the amortization period can be considered perpetual. A closed amortization period would identify a fixed date in the future, albeit up to 30 years, in which this liability is to be paid off.
For example, consider this estimate: if the plan were to adopt a 15-year funding strategy to amortize $11 billion of unfunded liabilities using the 8% interest rate and assuming pay increases at 4%, the total annual contribution (including the “normal” or ongoing cost) would increase from the current 9.39% of payroll to approximately 16% of payroll. This would represent an increase to taxpayers of about $500 million annually. Of significant note, such amounts do not reflect any recognition of asset losses since July 1, 2008which have reportedly been significant given the performance of the world’s financial markets.
SCRS Funding Considerations
The SCRS should refocus its entire funding strategy to make benefit costs current, affordable and predictable. Each of these terms is important. If such a standard cannot be achieved, a new system should be developed lest taxpayers both current and future are forced to face unaffordable tax increases.
To clarify terms:
Current costs are those in which the funding matches the accumulated obligations, based on the actuarial cost standard employed. These plans have chosen a commonly used actuarial cost method that attempts to establish a pattern of costs relatively constant as a percentage of payroll during the career of the employee. The goal should be to achieve a 100-percent ratio of assets to accrued liabilities (which are obligations earned to date). This is separate from, and in addition to, the annual cost of the value of benefits earned during the current year, commonly referred to as the normal cost.
Affordable costs are those in which total annual employer costs (net of employee contributions) fall within the range of 5% to 7% of payroll.
For employees opting for participation into the Optional Retirement Program (ORP) – a defined contribution option, taxpayers contribute 5% of payroll into individual accounts. This cost is consistent with the guideline described above.
However, the taxpayer is also required to contribute an additional amount into the SCRS to “equalize” the payment between the defined benefit and defined contribution plan, although the ORP participant receives no additional benefit from this additional contribution. The logic of this practice should be re-examined, as it offers no real benefit to either the public or participants.
Predictable costs are those constituting a relatively stable percentage of payroll.
The taxpayers of South Carolina are facing significant unfunded liabilities in their public pension and retiree medical plans. Current unfunded liabilities of SCRS and in retiree medical plans are approximately $20 billion with additional liabilities either deferred or yet to be quantified, especially those relating to the recent economic downturn. Based upon current funding practices, these costs will become a significant burden for future taxpayers. Actions need to be taken sooner rather than later to both responsibly manage these liabilities and fund these plans in a manner consistent with the pattern in which benefits are earned.
There are some measures SCRS should take to better protect the plan in the future. Some suggested actions include:
Conducting a thorough actuarial analysis to explicitly quantify the additional liabilities associated with the TERI program
Designing the retirement program to achieve cost profiles which are current, affordable and predictable – specific terms defined in this report. This would include reducing the amortization period of the unfunded liability to 15 years and achieving a long-term employer cost of 5% to 7% of payroll. If the existing plan cannot meet these objectives then provisions, including but not limited, to TERI and pension COLAs should be reduced or eliminated.
Implementing funding policies and actuarial assumptions that do not impose unnecessary risks and additional deferred costs on future taxpayers. In that context, the higher investment assumption and the 10-year asset averaging should be re-evaluated to minimize those risks.
Benchmarking benefit plans to include employer practices from South Carolina’s private sector.
For Retiree Medical Plans – GASB 45 Liabilities
Benchmarking benefit plans to include employer practices from South Carolina’s private sector.
Adopting funding policies generally similar to pensions which do not defer significant amounts of liability while being funded at a level consistent with taxpayer’s ability to pay.
About the Author
|Richard C. Dreyfuss is a business consultant and actuary with extensive involvement in compensation and employee-benefit plans, and broad experience in human resources and process management. He is also a senior fellow with the Commonwealth Foundation in Harrisburg,Pa., an independent, non-profit research and educational institute that develops and advances public policies based upon free-market principles. Dreyfuss worked for The Hershey Company (formerly Hershey Foods Corporation) for 21 years, and held numerous positions there, including director of compensation and benefits, prior to his retirement in 2002. He continues to be active in public policy matters, having testified in Washington, D.C., and Harrisburg, Pa., and was chair of The Pennsylvania Health Care Cost Containment Council in 2001–02. His emphasis is in areas that include process reengineering, human resources management, advocating policies to reform the pension and health care systems, and strategies to responsibly manage long-term employee benefit costs. Dreyfuss is a frequent speaker, and has been quoted widely in publications including The Wall Street Journal and newspapers throughout Pennsylvania. He is a graduate of Connecticut College, with a BA in Mathematics and Economics, and he earned a master’s degree in actuarial science from Northeastern University.|
The current actuarial valuation describes the group as public school employee, public higher education personnel, state employees, and city, county and other local public employees.
Non-active members include TERI participants, retirees & beneficiaries and inactive members.
 Beginning with the July 1, 2008 actuarial valuation, the actuarial value of assets is equal to the market value of assets less a ten-year phase in of the excess (shortfall) between expected market investment return (including the return on TERI balances) and actual net investment income (amounts determined prior to July 1, 2008 remain with a five-year phase in) with the resulting value not being less than 80% or more than 120% of the market value of assets.
 The actuarial liability is the difference between the total present value of future benefits and the actuarial present value of future normal costs. In layman’s terms, it is the present value of benefits earned to date by actives and non-active members.