In the second part of our series on city credit quality, the spotlight shifts to the Achilles heel for many cities: public pensions.
by Richard A. Ciccarone
Governmental credit quality is a byproduct of both the underlying economy and the cumulative decisions made by officials and citizens over time. The potential for severe strain tends to increase when both the economy and fiscal management break down, which can become even more likely if huge governmental liabilities loom in the backdrop.
The slow, grinding recovery from the 2008 credit crisis has helped most state and local governments restore their coffers through an uptick in revenues. On the other hand, a number of cities remain hampered and exposed to a host of significant liabilities, especially related to retirement benefits.
Using fiscal 2013 year data drawn primarily from city Comprehensive Annual Financial Reports (CAFR), Merritt Research Services, LLC examined a number of key ratios and statistics to ascertain the overall credit quality of America’s cities.
In this second part of our series, “Assessing the Credit Quality of America’s Cities,” we place the spotlight on the Achilles heel for many cities: public pensions. Future installments of this series will focus on liabilities encompassing other forms of debt, including other post-employment benefits and the average age of infrastructure, a liability whose time will eventually require funding. Wrapping up the fiscal assessment picture, we will examine overall condition measures such as deficits, financial cushions and net position.
How Adequately Are Today’s Public Pensions Funded?
For the last several years, significant attention has been focused on pension underfunding – specifically, the risk that retirement costs for past services pose to funding current essential services and infrastructure demands. The dramatic fall in the markets in 2009 made clear the folly in relying on optimistic assumptions, understated pension cost estimates and inadequate funding that saddled many cities as well as states with insufficient resources to cover them.
While underfunded states and territories, like Illinois, New Jersey, Rhode Island, California and Puerto Rico received most of the early attention, pension burdens in municipalities such as Vallejo, Stockton, San Bernardino, Central Falls, RI and Chicago, among others, provided abundant evidence that pensions pose threats to fiscal stability for many local governments, too.
Still, recent indications and trends are not all gloom and doom – at least not for most cities.
Based on the most recently available pension fund data found in the hundreds of cities reviewed in the study. Comprehensive Annual Financial Reports (CAFR) examined in this study, there’s a modicum of good news. Pension funding adequacy appears to be stabilizing for most cities since the low point in 2010. Although it is too early to say that the worst is over, positive market returns, along with a host of pension reforms, are producing some stabilization in the pension funding numbers for most cities.
The median combined funding ratio for all cities that have at least one single-employer pension plan has hovered around 75% for each of the last three years (2011-2013). That’s just a tad better than the low point for city pension plans, which occurred in 2010, when the median city pension funding hit 74%, but still well below the median pension funding ratio for cities that existed prior to the credit crisis in 2006 when better market returns raised expectations for pension assets and the median city funding ratio come close to 82%.
While trend information is helpful to assess the current directional condition of pension plans, medians can divert attention away from individual cities that are struggling with rising pension costs and serious underfunding as well as from cities that have prudently managed their own plans.
A number of governments have been able to achieve strong pension funding ratios primarily because they have responsibly maintained their stated plan assumptions as well as employer contributions. For example, Troy, MI, an outer suburb of Detroit, has shown a 90% or better combined pension funding ratio for the last several years, despite the fact that its pension plans use a conservative discount rate of 6.5% (The lower the discount rate, the more likely a pension plan will meet its investment projections, which in turn enhances the validity and credibility of a high funding ratio).
Still, perfection doesn’t come easily. A twist in Troy’s case is that its combined funding ratio of 95% would have reached 100% if it were not for the fact that the city’s much smaller Volunteer Firefighters plan is highly underfunded with a 38% funding ratio for 2014.
Public pension plans generally fall into three types of which the first two are similarly aligned relative to accountability:
- “Single-Employer“ and “Agent Multiple-Employer“ pension plans. Under a single-employer plan, the most common type of public pension plan offered by cities today, the employer (the city) is responsible for providing resources sufficient to make payments to plan participants. Similarly, an agent multiple-employer plan is one in which the assets of the participating government employers are pooled for investment purposes but separate accounts are maintained for each individual employer and responsibility remains with the participating city.
- “Cost-Sharing Multiple-Employer“ pension plans. As defined by the GASB, a cost-sharing multiple-employer plan is “one in which the participating government employers pool their assets and their obligations to provide defined benefit pensions-meaning that plan assets can be used to pay the pensions of the retirees of any participating employer.”
Single-Employer Public Pension Plans: A Mixed Funding Bag
In examining the 2012 and 2013 Merritt Research data relating to single-employer and agent pension plans, we found a fairly even number of cities whose combined pension plans are in strong condition (90% funding or higher) as cities whose plans are in what we would deem poor condition (less than 60% funding with a minimum of over three million dollars in unfunded liabilities).
According to combined single-employer and agent city pension reports for 2012 (derived from a February 2015 Merritt Research analysis of 1,112 cities), the specific breakdown by funding range was as follows:
- 15% achieved a favorable funding level of 90% or greater;
- 52% were funded in the 70% to 89% in the “getting by” range;
- 15% were in marginal condition with funding ratios between 60 and 69%; and
- 15% were in highly unsatisfactory funding condition with assets comprising less than 60% of liabilities.
- Note: 3% of the sample was comprised of cities with single-employer plans that also had less than a 60% funding, but a total unfunded actuarial liability of less than $3 million.
Pension reporting found in a city’s CAFR often lags a city’s fiscal year financial results by one year. Consequently, our February 2015 analysis of 2013 results was conducted with a somewhat smaller sample of 700 cities. Nevertheless, the findings regarding the breakdown by funding level remain consistent. The interesting, if not significant, difference between the two sets of data is that the most favorable outlier funding tier each grew by 1% while the weakest tier stayed relatively flat.
Single-Employer City Pension Plans to Monitor
The weakest-tier funding plans with ratios under 60% normally pose the most obvious and serious future challenge to spending priorities and the most risk to taxpayers. These plans will require not only continued scrutiny and angst, more than likely; they will need either or both increased funding and reform in order to safeguard pensions for their retirees as well as active workers.
Among the large cities reporting weakest 2013 funding ratios (in parenthesis) are: Portland, OR (1%);Springfield, MA (29%); Chicago, IL (34%); Louisville-Jefferson County, KY (36%); Philadelphia, PA (47%); New Orleans, LA (45%); New Haven, CT (46%); Tucson, AZ (56%) and Pittsburgh, PA (58%).
While Portland, OR stands with only a 1% funding ratio, its situation is certainly unique. Moody’s, the one rating agency that evaluates Portland, doesn’t seem worried, since it assigns it a top grade of “Aaa.” While most of Portland’s workers are currently enrolled in a state multi-employer plan, the 1% funding ratio relates to its Police & Fire Disability and Retirement plan.
In 2007, the City’s voters decided to make changes to the existing plan, funding it on a pay-as-you-go basis funded by an annual levy of a “capped” property tax. This approach has its own downside risk over the long term, especially if its tax base ever experiences a sharp, unexpected decline while the retirees are still being covered by the dedicated property tax.
Multiple-Employer City Pension Programs: Often Overlooked in the Headlines
Most of the headlines voicing concern with city pension problems focus on the single-employer plans and agent plans. Often overlooked are the local retirement programs that are part of a cost-sharing multiple-employer program, usually run by a state entity. We estimate that close to 40% of all cities contribute to at least one pooled multiple-employer cost-sharing pension program.
Occasionally, cities like New York, maintain both single-employer and multi-employer plans. While multi-employer managed local pension funds don’t always receive the same scrutiny as cities with their own pension funds, their burden and long-term liability risk on local taxpayers and city budgets can be just as challenging, especially if the multi-employer program isn’t properly funded or managed. Multiple-employer cost-sharing plans in the private sector involving corporations have actually experienced deficiencies when one or more employers is unable to fund its payments.
One relevant question in comparing single-employer plans and multiple-employer plans is: Are cities involved in multi-employer plans any better off related to the amount that they must contribute to their plans than their peers who offer single-employer plans? Do the contributions they need to make to these plans represent a smaller share of either the general fund or of all governmental activities accounts? Our analysis of the data found that cities involved in the multi-employer approach are spending slightly less relative to their governmental budgets than cities that run their own plans. However, the degree of difference is generally immaterial and not conclusive of a preferred approach. Likewise, many cities involved in the multi-employer plans spend a similar proportion of their budgets on their actuarial share of a multi-employer plan than cities that have responsibility solely for their own plans.
Based on 2013 year data derived from 653 cities reporting at least some participation in state multi-employer plans, the median spent on pensions relative to the general fund was 6.4%. These contribution levels are slightly lower than the median of around 7.6% of the general fund for 1,085 cities that rely on their own plans to pay pensions.
Using the general fund as a benchmark for expenditures doesn’t always present a fair apples-to-apples comparison to measuring the true budget allocation burden for cities from different states since state and municipal statutes may vary as to what program expenses are contained within the general fund versus a special revenue fund, capital fund or any another governmental function fund. For instance, some states may include a large portion of the public safety or road fund in a special revenue account, which makes the general fund sizably smaller than it would otherwise be.
In order to account for the mismatch, a comparison can be used to count pension payments made in a single year against all governmental activities accounts, which includes the general fund, special revenue funds, the capital improvement as well as debt service funds. By using the broader denominator of governmental activities expenditures, multi-employer pension plans still show a slightly lower spending share than the single-employer plans against the general fund, but the absolute burden percentage falls because of the wider base of total expenditures.
Based on our study, cities enrolled in multi-employer programs had median pension payments that amounted to 4.3% of their total governmental activities expenditures compared to a median burden of 5.3% for cities solely responsible for their own plans.
The number of large cities participating in multiple-employer pension plans that are carrying a fairly significant annual commitment to pay for pension contributions is comparable to the number in the single-employer group. Using “total governmental expenditures” as the benchmark comparison measure, which better equalizes the variation in what comprises a governmental account beyond the general fund, we found that a number of cities spent at least twice the 4.3% median related to participants in multi-employer plans.
The following list includes examples of cities making higher payments in 2013 relative to their total governmental expenses than the year’s all-city median (5.3%): Henderson and North Las Vegas, NV (both 12.8%); Colorado Springs, CO (12%); Santa Fe, NM (11.6%); Albany, NY (11.3%); Louisville-Jefferson County, KY (10.9%); Reno, NV (10.2%); Madison, WI (10%); Columbus, OH (9.4%) and Cleveland, OH (8.8%).
Interestingly, New York City, a hybrid situation, had a combined 2013 pension burden of 11.6% of its total governmental expenditures. By comparison with another large city (one that only participates in a single-employer pension plan), Los Angeles’ actual contributions to its pension plans amounted to 15.3% of its total governmental expenditures.
Then there are numbers behind the numbers. Chicago’s 2013 pension payment appears to be relatively manageable against both all of its governmental activities expenses (6.3%) as well as the general fund (14.3%). However, Chicago’s payments would have soared much higher to about 25% of its governmental activities accounts for the same year had the city contributed the full amount of its actuarial required contribution (ARC) or annual pension cost (APC), which is the ARC plus certain amounts not paid in the past under existing GASB rules.
Boston appears to be notable on the low side of pension spenders among big cities. Its contribution to its State-Boston Retirement multi-employer pension program in 2013 amounted to only 4.7% of its governmental activities expenses and just a little more than 5% of its general fund. However, before any accolades are awarded to Boston, it should be mentioned that the overall funding ratio of this multi-employer plan is only in the low 60% area.
For that reason, Boston’s lower pension spending today could be a reverse harbinger of larger payments down the road. The optimum position would be to have low contribution payment record associated with a highly funded, conservative assumption pension plan.
The slight advantage to cities participating in a state-administered multi-employer pension programs is most likely a product of improved economies of scale and a more rigorous employer pension actuarial contribution. In contrast, the Merritt database shows the 2012 median city single-employer plan contribution to be nearly 100% of the actuarial required contribution; however, nearly 30% of all cities (single-employer and multi-employer plans) allocated less than the full requirement.
One other factor helping multi-employer plans is that the pool of assets to be invested is larger than what would be managed for any single city, which enables a more sophisticated investment platform by lowering fees, spreading the risk and effectively broadening the potential for some higher investment returns through somewhat more aggressive strategies.
Pension Fund Discount Rates – A Key Factor to Getting the Actuarial Assumptions Right
There are multiple reasons that pension underfunding has become a significant problem for many cities. On paper, pension funds should be in a good position to pay 100% of the future benefit payments they owe to their retirees as long as all scheduled actuarial contributions are made, investment returns meet projected results and payouts to retirees are in line with forecasts. Unfortunately, the theory behind actuarial forecasts has proven to be easier said than done as assumptions used to predict the future often prove to be mistaken.
The most common assumptions that have often diverged from forecasts include: mortality rates, retirement ages, inflation expectations, back end salaries, inadequate governmental contributions relative to the actuarial requirement and overly optimistic investment returns. The latter factor has been the primary object of discussion and criticism particularly in the aftermath of the market swoon which took place during the credit crisis.
Under the GASB accounting rules, the discount rate is used to calculate the future actuarial value of whether a government has sufficient assets to cover its liabilities based on expected investment returns. Therefore, as the discount rate goes higher, the government can justify lower contributions to the plan under the assumption that it will eventually cover all promised future pension payments to retirees as well as active workers since the fund is earning investment monies in lieu of monies that that would otherwise be required by either the government or workers to fully fund the plan.
Market and public criticism that public pension discounts have been too high has led many governments to begin lowering them especially since the Great Recession. Still, the median discount rate of all city pension plans has only marginally fallen since 2009. The median discount rate for the approximately 1,500 city pension plans included in the Merritt Research study covering the period 2010-2012 stood steadfast at 7.5%, only a slight decline from 7.75% in 2009. For the year 2013, using a smaller sample of 800 pension plans, the discount rate remains unmoved at 7.5%.
Among city pension plans reporting 2013 numbers, a breakdown of discount rates follows:
- 22% still had discount rates of 8% or higher;
- 32% were in the 7.5% to 7.95% range;
- 32% were in the 7% to 7.4% range; and
- 14% were more conservative with discount rates lower than 7%.
Based on 2013 pension reports, the cities of Gainesville, FL (8.5%); New Haven, CT (8.25%), and Springfield, MA (8.13%) are among the cities that are most optimistic about their investment returns, with at least one single-employer plan with discount rates over 8% (shown in parenthesis).
How the New GASB Rules Impact Public Pension Reporting
The GASB has adopted new pension reporting rules applying to cities, effective after June 15, 2013 for Statement 67 and after June 15, 2014 for Statement 68. Some of the most significant changes to be implemented are: the elimination of multi-year smoothing (averaging) of investment assets, new limits on amortization periods in line with service periods of workers, more standardized reporting on multi-employer plans that require proportionate share pension liabilities to be reported for participants, the elimination of the Actuarial Required Contribution (ARC) determination, broadening the full pension liability that will apply to the city’s net position balance sheet and the potential for a blended discount rate tied to investment assumptions.
These changes are likely to have both positive and negative impacts on funding ratios. Although the essence of the new GASB approach will be to move away from using funding principles to an accounting approach, analysts will be quick to assess the numbers to gauge whether or not the government is fulfilling its responsibilities to have sufficient assets to cover liabilities.
The elimination of asset valuation smoothing will make funding ratios more volatile on a year-to-year basis and have the immediate effect of improving funding adequacy ratios to reflect positive market returns in good years and lower funding ratios in bad years. Shorter amortization levels will require higher funding requirements near term. Standardization in the multi-employer plans will make those plans more comparable to one another. New rules for discount rates should be interesting to monitor.
A much lower discount rate to the current investment return assumption will be required to be used whenever the actuarial study shows that the plan will become deficient in its ability to pay its active workers and retirees. At that point, all unfunded liabilities beyond a so-called crossover point will have to use a discount rate equivalent to an “AA” or higher tax-exempt bond or index, which recently, has been running at or below 4%. Those required to move to the lower blended discount rate will see a proportional reduction in the funding ratio that they had been using prior to the implementation of the new GASB rules.
Until recently, the threat of a two-tiered discount rule suggested that there might be a relatively widespread initial drop in public pension funding ratios; however, recent evidence from early reporting governments applying to 2014 pension reports suggest otherwise.
One of the early reporters to the new accounting rules is the New York City Police Pension Fund. In FY 2012, the plan reported a 61% funding ratio, so it was interesting to see how the City’s net position would fare under the new discount rules. The results appear to be instructive and, for some cities, a surprise to the upside. Under the new procedures, New York City Police Pension Fund’s revised funding ratio, in accordance with the new rules, rose to 66% in 2013, markedly higher than for FY 2012. Further reinforcing the positive impact, the City’s 2014 pension funding ratio escalated to 74%. The New York City Fire Pension Fund experienced a similar uptrend.
Probably the most important reason that New York City’s net pension position improved by so much in one year was the fact that its asset revaluation in light of the overall positive market returns on its assets, weren’t constrained by a “smoothing” effect which would have otherwise averaged out returns over a longer time period.
Year-to-year market impacts, as was shown in New York, are more likely to have a substantial, immediate effect on funding ratios. But, on the downside, funding ratios are more likely to be far more volatile. The New York City Police Pension example may be particularly enlightening because it was not required by their actuaries and accountants to apply a lower blended discount rate since their actuarial analysis did not determine that the plans would reach a crossover point that projected insufficient funds to cover retiree and active worker pension payments.
Based on the early reporting example of New York’s financials and pension plan reports, we would expect that the potential negative impact of the lower blended discount rate is likely to be triggered in only a minority of public pension plans. More than likely, all other factors being equal, we expect that funding ratios for cities are more likely to improve for state and local pension plans as the 2014 and 2015 results are released as long as investment returns remain positive.
An important caveat here is that since market volatility will be an important factor in determining asset sufficiency, a low funded plan could float between the traditional investment rate of return and the more punitive blended rate on a year to year basis.
Lastly, the new rules will more than likely increase the net pension liability for nearly every city. Under the old accounting rules, governments only accounted on their GASB 34 balance sheet a portion of the pension liability linked to a period synchronized with the effective date of the older pension rules.
This “unfunded pension” amount was labeled as the net pension obligation and it was shown as a liability on the net asset governmental activities balance sheet of the city. Going forward with the new GASB rules, the full amount of the pension liability will be applied to the city’s bottom line net position.
Exploring the Parallel between Pension Ratios and Bond Ratings
How closely associated are ratings to pension funding ratios? The Merritt Research data found mostly a positive correlation between funding ratios and bond rating levels for cities (based on our analysis done in October 2014 plans involving over 1,400 cities using 2012 data); however, the relationship was weak once a city was related in the lower rating grades.
Using the highest rating assigned any of the three agencies for the Merritt rating groupings, those cities, which had at least one AAA rating recorded the best group median funding ratio (76.1%), followed by those cities with at least one AA rating (74.7%) and then those cities with their highest rating at the A rating (69.4%). But then, the positive correlation breaks down below the single A level.
At that point, the median pension funding ratios for cities goes up and flattens out for cities in which their highest rating is in the BBB bracket (72.3%) or below BB (72.9%), and single B (72.8%). These groupings suggest that pension funding ratios may be a factor in rating assignment; but, by no means is this factor a sole determinant. From a statistical standpoint, the much smaller number of cities rated below A by the agencies tends to decrease the correlation significance of outlier numbers.
Richard A. Ciccarone is a Co-Publisher of MuniNet Guide. He is also President and Chief Executive Officer of Merritt Research Services, a municipal bond credit database and research company that primarily serves institutional investors, investment dealers and bankers.
Coming soon: Assessing the Credit Quality of America’s Cities: Part Three, which will focus largely on long-term liabilities other than pensions.
Did you see Part One of this series: Focusing on Bond Ratings?