Detroit's proposed pension reform from last week has a chance to be precedent setting. The city's plan to "adjust" its $18 billion in debt by slashing as much as half of it before exiting bankruptcy includes cuts to already accrued pension benefits, as well as changes to how future public pension benefits are earned. The proposal also has less innovative, but important elements of good pension reform including forcing the city to use "market values" in accounting for its liabilities, using slightly more realistic investment return assumptions, and adjusting its defined-benefit system.
Such structural changes are bound to bring criticism. Retirees have already disseminated their displeasure with the proposed pension reform. The Official Committee of Retirees claims that the proposed changes would push 20% of the city retirees into poverty, and argue that the accounting changes are overly conservative. Other retirees have asked why the city hasn't sold the pieces at the Detroit Institute of Arts in an auction, or sell Belle Isle, in order to lower the need for pension cuts.
These retires should be more concerned that the Detroit pension reform proposal in the bankruptcy plan doesn't go far enough in its changes to protect their pension benefits. The city has retained its defined-benefit structure for future earners and is proposing to have retirees bear the risk of future accrued liabilities. This means employees would have all of the risks of a defined-contribution plan without the benefits. And while representatives of the retirees are claiming the bankruptcy plan's pension reforms are too conservative, they are probably not conservative enough.
The bankruptcy plan also includes a "DIA Settlement": $365 million in donated money from a group of chartable foundations to prevent pieces at the Detroit Institute of Arts (DIA) from being sold to limit pension cuts, a promise from Gov. Snyder that the state will contribute $350 million to the pension fund, and a promise from the DIA to raise $100 million from donors over 20 years for pension funding. This idea has problems of its own, but won't be discussed here beyond a note that the money promised from the foundations is almost certainly less than could be gained from a private auction and pensioners have reason to criticize the DIA Settlement approach.
The following analyzes the details of Emergency Manager Orr's proposed Detroit pension reform for the two city pension systems: the General Retirement System (GRS) for all non-uniformed city workers, and the Police and Firefighter Retirement System (PFRS). All quotes come from the bankruptcy plan "Disclosure Statement" and its appendices.
New Funding Ratio Target
Detroit aims to achieve a 70% and 75% funded status for GRS and PFRS, respectively each pension plan by June 30, 2023. Funding ratios will be based on the market value of assets, not a smoothed value of assets (Detroit currently determines its liability by "smoothing out gains and losses over a seven year period). -- See p.5 of "Disclosure Statement"
Analysis: These funding targets are still rather low to be considered a well-funded system, even on a market-value basis. GRS is presently 59% funded on a market-value basis and PFRS is 78% funded by the same measure. Given that Detroit should not have to go through bankruptcy pain again, the city should consider reducing future benefits more rather than have such underwhelming funding ratio targets.
Changes to Actuarial Assumptions
GRS will lower its assumed investment rate of return from 7.9% to 6.25%. PRFS will lower its assumed investment rate of return from 8% to 6.5%. The bankruptcy plan defends these changes as follows:
"The use of these investment return assumptions is consistent with the trend by governmental entities to reduce pension funding assumptions, and the particular rates used in the Plan - although lower than most jurisdictions - nonetheless align with the unique financial inability of the City to weather unanticipated pension investment loss. These conservative assumptions are also appropriate given the large percentage of investments held by the pension funds that do not have a readily determinable market value and the uncertainty to actual asset values held by the pension plans as a result." -- See p.6 of "Disclosure Statement"
Analysis: The new discount rates of 6.25% and 6.5% may be lower than most municipal assumed rates of return, however, that shouldn't be a cause for concern. Most assumptions are unrealistic, and these rates are at the top of what should probably be assumed.
Cuts to Retiree Benefits
GRS retirees and their beneficiaries will see a 34% reduction in their monthly pension. PFRS and their beneficiaries will see a 10% reduction in their monthly pension. And there will be no cost-of-living-adjustments (COLAs) for the next decade.
If the Detroit pension board agrees in a "timely manner" to the plan of adjustment, then the GRS retiree pension checks will only be cut 26% monthly, with no future COLAs. PFRS pension checks would only be cut 4% monthly, with no future COLAs. -- See p.7 of "Disclosure Statement"
Gov. Snyder has offered state funds as a part of the DIA Settlement in exchange for a guarantee that no retiree would see their benefits cut below the federal poverty line. The bankruptcy plan does not yet detail how this would work, but does say "additional benefits will be provided to [GRS and PFRS pensioners] who... have household income less than a threshold amount to be determined pursuant to further discussions between the City and the State and tied to federal poverty levels." -- See p.8 of "Disclosure Statement"
Analysis: It is a positive sign that the city is proposing cuts to both current and future benefits. This is unfortunately the only way that most benefits for retirees can be preserved while also balancing the need to use tax dollars for city services. Whether or not the state's bailout funds for the art are approved, the cuts to retiree benefits should certainly be means tested so that no retiree currently receiving a pension is pushed below the poverty line. However, even with these positive marks, the city should consider additional cuts in order to improve the targeted funding ratio.
Cuts to Active Worker Accrued Benefits
GRS retirees and their beneficiaries will see a 34% reduction in their monthly pension. PFRS and their beneficiaries will see a 10% reduction in their monthly pension. Plus, any additional unfunded liabilities added in either system during FY2014 will be also cut from accrued benefits. Also, for any active workers who retire in the next decade there will be no COLAs.
If the Detroit pension board agrees in a "timely manner" to the plan of adjustment, then the GRS accrued benefits will only be cut 26%, with no future COLAs. PFRS accrued benefits would only be cut 4%, with no future COLAs. -- See p.7 of "Disclosure Statement"
Analysis: Having active workers share in the cuts at the same level as the retirees is likely a good public relations step. Making active workers responsible for additional liabilities is also likely to be a good way to align incentives of all parties involved to ensure the system is properly funded.
From the initial reaction, the tacit bribe of lower cuts if retirees accept the plan without further litigation is going to be a no-go. However, Detroit retirees should consider the fate of other cities where retirees rejected reforms only to be forced into deeper cuts later on, such as Central Falls, RI.
Cuts to Future Benefits & New Risk Management
Future benefits for active workers will be earned via a new formula (the details are slightly different from GRS and PFRS, the later of which gets a better pension -- see p.18, 22 of "Disclosure Statement").
The future benefit formulas would be for defined-benefit pensions, like the old system, but at reduced rates. Active workers would therefore be put into a "hybrid" system, but one that is still completely defined-benefit (unlike hybrid defined-benefit/defined-contribution systems found in Indiana or Rhode Island).
The system contains rules that "shift funding risk to participants in the event of underfunding of hybrid pensions" and creates new minimum retirement ages for accrued benefits under the old system.
Analysis: This "hybrid" approach does not go far enough to reform Detroit's pension. The bankruptcy process is the perfect time to put Detroit on a path to ridding taxpayers completely of pension liabilities by freezing the defined-benefit system to new members. A defined-contribution system like a 401(k) would also be more portable for workers and enusre their benefits are not subjected to political underfunding of pensions.
Possibility of Restoring Benefit Cuts
If the pension systems recover at a greater rate than expected, benefit cuts can be restored under certain conditions: not until at least FY2024, only if the funding level of a system is above 80% based on market value of assets using a 6.25% discount rate for GRS and 6.5% discount rate for PFRS, and only if the pension board approves the restoration benefits. -- See p.7 of "Disclosure Statement" and "Exhibit A, Article I" items 138-143 & 192-198
Detroit will also be recalculating the allocation of resources to the GRS Annuity Fund and any determined "excess allocations" will also be available to restore some benefit cuts. -- See p.7 of "Disclosure Statement"
Analysis: This is a helpful public relations element of the plan, but the cuts are already likely not enough. Further, even if the systems did become better funded after a decade, the threshold for restoring some benefits should probably be 95% or 100% instead of 80% funding on a market-value basis
Changes to Contributions
The City of Detroit will save cash by not making any annual contributions to GRS or PFRS for the ten years after the plan of adjustment. As a replacement, pension funding will come from the DIA Settlement
Pension funding will come instead from the DIA settlement. The Detroit Water and Sewage system will also contribute $675 million to the GRS fund over ten years. After FY2024, Detroit will need to start making contributions to the pension funds again.
Analysis: This approach is primarily aimed at freeing up cash to invest in city infrastructure and services over the next decade. The idea is that public works projects to improve the city would help Detroit grow again, expanding the tax base so that there will be more available cash to make pension contributions down the road. But Detroit's challenges are because of aging infrastructure. Improving services, particularly security, would likely keep more people from leaving, but the real reason why the city isn't expanding has more to do with the tax and regulatory code.
Health Care Changes
The city would create a Voluntary Employees' Benefit Association to take over the health care liabilities of GRS and PFRS. This fund would be started with $526.5 million. It would replace all life insurance and death benefits. -- See "Exhibit I.A.182.a".
No Changes for a Decade
The bankruptcy plan also would require no changes to the terms or assumptions over the next decade.
The bankruptcy plan is explicitly using monies promised to the city through the DIA Settlement to lower the cuts needed for retiree pensions. If the foundations do not follow through with their committed contribution, then pension benefits will be cut by the same dollar amount. -- See p.6 of "Exhibit A"
These commitments will be spread out over 20 years. In exchange, any assets that the "City holds title to" would be considered transferred to DIA under the same "perpetual chartable trust" that the bulk of the art in the museum is held under, and would be required to remain "within the City limits, for the primary benefit of the residents of the City and the Counties and the citizens of the State." -- See p.98 of "Settlement Disclosure" and p.5 of "Exhibit A".
Gov. Snyder proposed $350 million state bailout for DIA would come from the state's tobacco settlement fund, but has not been approved yet by the state legislature.
Approximately $50 million can go to GRS before FY2024, and $195 million will be paid to GRS after FY2024. Approximately $175 million can go to PFRS, before FY2024. This would leave up to $395 million in proceeds from the DIA settlement without specific distribution outlined in the plan.
Is Reform Necessary?
In a recent post for Out of Control, I outlined the case that Detroit's pension systems are not as well funded as some claim. On the surface the numbers look okay, but the actuarial assumptions are hiding huge losses, which if reflected on the books would reveal a pension system that is not capable of paying out all of the benefits promised to retirees.
Opponents of the Orr bankruptcy plan proposal are acting rationally, but only based on their presumption that the pensions are in fine shape. If, on the other hand, the systems are not well funded than opponents of reform are not acting in the interests of pensioners who have been promised benefit distributions 10 and 20 years from now.
The bankruptcy proposal includes a discussion of how the funds are in trouble, and concurs with our analysis from earlier in February. Here are some highlighted admissions in the plan:
"The City believes that the UAAL figures reported by the Retirement Systems were substantially understated because they were based upon various actuarial assumptions and methods that served to substantially understate the Retirement Systems' UAAL. The assumptions and methods included: (a) annual net rates of return on investments (GRS -7.9%; PFRS - 8.0%) that were unrealistic in light of the Retirement Systems' demographics, the targeted mix of the Retirement Systems' assets and the inability of the City to budget for and fund pension investment loss in the event the sought-after returns were not achieved; (b) the "smoothing" (reallocation over a period of years) of asset gains and losses over a seven year period, which masks the funding shortfall; and (c) the use of 29 year (PFRS) and 30-year (GRS) amortization periods for funding UAAL - which is applied anew each year to the full amount of unfunded liability - that allows unfunded liabilities to continue to grow rapidly as a result of compounding. In the List of Claims, the City set forth what it believes is a more realistic total UAAL for the Retirement Systems of $3.474 billion, consisting of $2.037 billion in UAAL owed to the GRS and $1.437 billion in UAAL owed to the PFRS... If one were to apply even more realistic assumed rates of return of 6.25% for GRS and 6.50% for PFRS, respectively, the UAAL totals increase to $2.299 billion for the GRS, and $1.588 billion for the PFRS, as of the end of Fiscal Year 2012." -- See p.5 of "Settlement Disclosure"
I would argue that even more realistic assumptions should be used, but even by these moderately conservative standards there are clear underfunding issues. Why should investment losses be spread out over seven years? Why should payments be amortized over such a long time period? Fixing these practices, which would highly suspect in private sector pension accounting (if not illegal), is necessary. And once that process begins it is clear that serious reforms are needed.
The case for reform is even stronger once you consider the suspect practices of the pension funds over the past decades. Even if the reform assumptions were overly conservative, that would not change the actual practices of the funds that have contributed to the low funding levels when measured at a market-value rate (instead of the funding-value level using poor actuarial assumptions). Consider the following list from the bankruptcy report:
"In the past, the Retirement Systems engaged in a variety of practices that contributed considerably to the underfunding of the pension plans, particularly with respect to the GRS pension plan. As more fully discussed in Section III.B.5, these practices included: (a) consuming pension fund assets to pay promised returns under the separate "annuity savings plan," (b) dissipating pension fund assets during the years when returns on investment exceeded expectations through the so-called "13th check" program, (c) deferring required pension fund contributions from the City each year and financing the deferred amounts at a rate of 8%. Serious allegations also have been made that various former officials of the Retirement Systems accepted bribes and/or misappropriated assets of the Retirement Systems for their own personal gain. In addition, the Retirement Systems have made many poor investments that have reduced the funded status of the two pension plans. Finally, it appears that a large portion of the assets of the respective Retirement Systems is invested in alternative investments for which no recognized market valuation exists. As of June 30, 2013, approximately 24% of PFRS assets and 33% of GRS assets had estimated, rather than readily ascertainable, market values." -- See p.5 of "Settlement Disclosure"
Combined, the history of Detroit's bad actuarial assumptions and its poor operating practices more than make the case for reform. And the Detroit plan of adjustment makes steps towards the needed reform, though there is more to do.