A Widening Gap in Cities

Shortfalls in Funding for Pensions and Retiree Health Care

Quick Summary

A number of the nation's key cities face unpaid bills for the retirement benefits they have promised their employees. Building on Pew's earlier work on state retirement systems, this report examines pension and retiree health care funding in 61 cities—the most populous one in each state plus all others with populations over 500,000—and looks at options for reform.

Executive Summary

As the Great Recession ended, 61 key cities across America—the most populous one in each state plus all others with more than 500,000 people—emerged with a gap of more than $217 billion between what they had promised their workers in pensions and retiree health care and what they had saved to pay that bill. While states have a much larger shortfall, cities face the same daunting challenges posed by unfunded liabilities for their public sector retirement benefits.[i] 


For pensions, these cities had a shortfall of $99 billion in fiscal year 2009, the most recent year with complete data. The gap continued to widen in fiscal year 2010 as reflected by complete data for 40 of the cities, which saw unfunded pension liabilities rise by another 15 percent.[ii] 

Besides pensions, many localities also have promised health care, life insurance, and other non-pension benefits to their retirees, but few have started saving to cover these long-term costs. These unfunded liabilities loom even larger than for pensions (see graphic).[iii] As of fiscal year 2009, the cities in this report had promised at least $118 billion more than they had in hand to cover retiree health care benefits.

Wide disparities exist in how prepared cities are to fulfill their pension obligations.  

  • Milwaukee and Washington, D.C., had surpluses at the end of fiscal year 2009, with enough money to cover 113 percent and 104 percent, respectively, of their liabilities, better than the best-funded state, New York, at 101 percent. [iv] 
  • In four cities—Charleston; Omaha; Portland, Oregon; and Providence, Rhode Island—pension systems were more poorly funded than those in Illinois, which at 51 percent was the lowest-funded state. [v] 
  • Charleston trailed all the other cities at 24 percent.[vi] 
  • Overall, the cities had enough money to cover 74 percent of their pension obligations in fiscal year 2009, compared with 78 percent for states.[vii] 

Cities have more in common when it comes to gaps in funding for retiree health care and other non-pension benefits.

  • As of fiscal year 2009, overall, cities had set aside enough money to cover just 6 percent of their promises, compared with slightly more than 5 percent in states.[vii] 
  • Only Los Angeles and Denver had even half of the money needed.

Report Contents

Key Findings: Pensions

  • The big picture. Altogether, the 61 cities had enough assets to cover 74 percent of $385 billion in projected pension obligations as of fiscal year 2009, the most recent year with complete data for each city. That left a gap between assets and liabilities of $99 billion. More recent, complete data for 40 cities showed the gap widening in fiscal year 2010.
  • Performance on two key indicators. Cities were assessed on: (1) their funding level, which is the percentage of projected liabilities covered by assets; and (2) the extent to which they are paying the annual contribution their actuaries recommend to meet their pension obligations, generally over 30 years.[i]   


    Between 2007 and 2009—one of the most volatile financial times in their history —16 cities maintained funding levels above 80 percent and consistently made at least 90 percent of their annual pension payments: Albuquerque; Baltimore; Charlotte; Dallas; Denver; Des Moines; Los Angeles; Milwaukee; Salt Lake City; San Antonio; San Francisco; Seattle; Sioux Falls; Virginia Beach; Washington, D.C.; and Wichita.
    Nine cities fell below the benchmarks for both funding level and annual pension contributions each year from 2007 to 2009: Charleston; Chicago; Fargo; Jackson; Little Rock; New Orleans; Omaha; Philadelphia; and Portland, Oregon.
  • Best- and worst-funded in 2009. Twenty-four cities emerged from the recession with funding levels of 80 percent or higher in fiscal year 2009; 37 cities fell below that mark.[ii] This snapshot captures cities' pension holdings at a low point because of the recession.
  • Best and worst at making annual payments. More than half—35—of the 61 cities made at least 90 percent of their annual pension payments each year between 2007 and 2009, including 25 that paid at least 100 percent each year. Six cities regularly shortchanged their pension funds and made less than two-thirds of their annual recommended contributions: Charleston, Chicago, Little Rock, New Orleans, Omaha, and Portland, Oregon.[iii] 
  • Results in 2010 and beyond. A look at 40 cities that reported results for all of their pension plans for fiscal years 2005 through 2010 showed declines in funding levels beyond the end of the recession. Their aggregate funding levels reached 82 percent in 2007, before dropping to 78 percent in 2008, 73 percent in 2009, and 70 percent in 2010.[iv]

Key Findings: Retiree Health Care

  • The big picture. Pension shortfalls grab more headlines, but they are not the biggest retirement bill coming due for the 61 cities. The cities had set aside just 6 percent of $126.2 billion in projected costs for what are known as Other Post-Employment Benefits, primarily retiree health care, leaving $118.2 billion in unfunded liabilities in fiscal year 2009.  

Many cities are just beginning to set aside contributions for these benefits, as they routinely do for pensions. Because retiree health promises generally have fewer legal protections than those for pensions, some cities are trimming them to reduce their long-term liabilities.

  • Best- and worst-funded. Los Angeles led the 61 cities with 55 percent of its retiree health care promises pre-funded in fiscal year 2009. Next were Denver (51 percent); Washington, D.C. (49 percent); Louisville (40 percent); Sioux Falls (37 percent); and San Antonio (31 percent). 

While 27 cities had set aside some assets to offset their liabilities, 33 had not and were paying for their retirees' health care out of their treasuries on a pay-as-you-go basis. One city—Portland, Maine—had no liabilities because it offers no retiree health care.

  • Best at making annual payments. Some cities have begun tackling their retiree health care liabilities by pre-funding a portion of their expenses. Of the 27 cities with some assets set aside, five contributed more than 90 percent of the annual sums recommended by their actuaries in both fiscal years 2009 and 2010: Anchorage, Charlotte, Los Angeles, Sioux Falls, and Virginia Beach. Eight contributed more than half of their full annual payments in both years, and 14 contributed less than half. 
  • Bigger unpaid bills for retiree health care than for pensions. In total dollars, the 61 cities had promised three times more in pension benefits than in retiree health benefits.[i] Yet more than a third of the cities faced bigger unpaid bills for retiree health care than for pensions.[ii] That is primarily because cities have saved for pensions for years, so a greater portion of those liabilities were covered.  

Lessons From the Recession

The Great Recession had a significant impact on local pension plans across the country. Overall, the aggregate funding level of the 61 cities studied declined five percentage points—from 79 percent in fiscal year 2007 to 74 percent in fiscal year 2009. Half of the cities saw drops of eight percentage points or more. But the downturn was not the decisive factor that separated cities with the best-funded pension systems from those with poorly funded ones. 

Whether a city was fiscally disciplined made a big difference in how it fared. Cities with pension plans that kept up with their payments—consistently making the “annual recommended contribution” calculated by their actuaries—weathered the financial downturn better than their counterparts.[i] Between 2007 and 2009, 35 cities paid at least 90 percent of each year's annual recommended sum. The funding level of their pension plans fell at half the rate as those in cities that did not consistently make the bulk of their payments.[ii] 

Cities began to reform their public sector retirement systems before 2007, but changes have accelerated rapidly in the wake of the Great Recession. Many of the 61 cities in this analysis—even those with relatively well-funded systems—have made adjustments to address funding shortfalls or unsustainable growth in costs for pensions or retiree health care.

In general, reforms for both pensions and retiree health fall into four categories: (1) plan design; (2) funding; (3) benefits; and (4) organization and management.

The changes most commonly affect new hires but also current retirees and employees in some cases. 

Pension and Retiree Health Care Funding in 61 Cities, FY 2007-2010


Executive Summary

 [i] This report is the broadest look to date specifically at cities' retirement liabilities, though it covers a subset of all cities. By comparison, the 50 states faced a gap between assets and liabilities of $1.26 trillion for fiscal year 2009: $660 billion for pensions and $604 billion for retiree health benefits, according to the Pew Center on the States report “The Widening Gap: The Great Recession's Impact on State Pension and Retiree Health Care Costs,” (April 2011), The gap grew to $1.38 trillion in fiscal year 2010: $757 billion for pensions and $627 billion for retiree health care, according to Pew's, “The Widening Gap Update,” (June 2012),

 [ii] Complete pension data for fiscal year 2010 were available only for 40 of the 61 cities. Pew used the most recent comprehensive annual financial reports available for cities during a final data collection period that began in March 2012 and lasted for several weeks. Even among cities that had issued their financial reports for fiscal year 2011, some reports lacked up-to-date data for one or more of their workers' pension plans. Pension data often are not as current as other financial information. Moreover, according to Pew's methodology, even pension valuations dated January 1, 2011, for example, would be counted as fiscal year 2010 data to reflect performance for the just-completed fiscal year. 

 [iii] This report uses “retiree health care” when referring to Other Post-Employment Benefits, a category that includes other non-pension benefits but in which costs are primarily from retiree health care. 

 [iv] The funding level for Washington, D.C., represents only liabilities accrued since 1997 in defined benefit pension plans for firefighters, police, and teachers. The federal government in 1997 took over financial responsibility for benefits accrued by those workers up to then, as well as retirement benefits for judges, relieving the city of $4.9 billion in unfunded liabilities that it had inherited from the federal government. The city also contributes to a defined benefit plan (for 2,700 general municipal workers hired before 1987) that is managed by the federal Civil Service Retirement System, but an estimate of the city's share of those liabilities could not be obtained. This study does not include the costs of pension benefits for general municipal employees hired since October 1987. They are not in a defined benefit pension plan but instead receive benefits through a defined contribution system in which the city deposits money each year into a retirement account for each employee. See Government of the District of Columbia, “Comprehensive Annual Financial Report 2011,” (September 30, 2011), 111–117,

 [v] Both Charleston and Portland, Oregon, can blame extremely low savings in police and fire pension plans for the city's low overall funding level. Charleston faced a shortfall of $193 million in its police and firefighters' pension funds, which were only 11 percent and 6 percent funded, respectively, in fiscal year 2009. Meanwhile, its larger general employees' fund was 80 percent funded, with a shortfall of $12 million.

Portland, Oregon, had virtually no assets to offset unfunded liabilities of $2.3 billion in fiscal year 2009 in its pension and disability plan for police and firefighters hired before 2007. Meanwhile, the plan covering the rest of Portland's employees was estimated at 86 percent funded, with unfunded liabilities of $453 million. Portland funds the retirement costs of police and firefighters hired before 2007 on a pay-as-you-go basis, meaning the city relies on property taxes each year to pay benefits and does not attempt to set aside money for the future to meet those liabilities

 [vi] States' pension funds were 78 percent funded in fiscal year 2009 and 75 percent funded in fiscal year 2010, according to Pew's “The Widening Gap” and “The Widening Gap Update” reports.

Key Findings: Pensions

 [i] In comparing funding levels, it should be noted that pension plans derive their funding levels based on different assumptions, such as for projected investment earnings or expected life span of retirees, that may vary from plan to plan and from city to city. Accounting methods also can vary.

 [ii] While the target funding level for a pension plan should be 100 percent, funding below 80 percent of actuarial accrued liability is widely recognized as inadequate. An 80 percent benchmark is used by the Pension Protection Act of 2006, a federal law that pertains to private sector pension funds, to determine when stricter funding rules apply. In the context of public sector pensions, 80 percent provides a useful dividing line for comparing pension systems. But there is debate about what funding level short of 100 percent signifies a healthy plan. For a discussion of this issue, see American Academy of Actuaries, “Issue Brief,” (July 2012),; or an explanation by pension experts Keith Brainard and Paul Zorn, (January 2012),

 [iii] Four of these cities actually made all or most of their annual payments for certain pension plans, but still fell short overall because they did not sock away much money for others. For example, Charleston and Portland, Oregon, consistently paid all or most of their annual recommended contributions for their general employees' pensions but not for those of police and firefighters. Portland does not even attempt to set aside money for the pensions of police and firefighters hired before 2007 but relies on property taxes to cover those pension checks each year, leaving unfunded liabilities on the books.

Over this time period, Little Rock consistently made 100 percent of the annual payments recommended to fully fund its long-term pension promises to police and firefighters hired since 1983. Its payment record overall appears much lower because Pew's calculation also covers three defined benefit plans closed to employees more than three decades ago. In such situations, a city may pay benefit costs as they arise, an amount often less than 100 percent of what is actuarially recommended. Pew's calculation does not take into account the city's annual payments to its defined contribution retirement plan, which has covered non-uniformed city employees since 1981.

In similar fashion, New Orleans made at least 80 percent on average of its annual recommended contributions to three current defined benefit pension plans for workers, while it did not make any contributions to two smaller closed plans for police and firefighters. 

 [iv] This result tracks trends from other studies, which show funding levels for different groups of public pension plans continued to decline in 2011. At the end of fiscal year 2010, the aggregate funding level for pension systems in the “Public Fund Survey” was 77 percent, down from nearly 80 percent in fiscal year 2009. The “Public Fund Survey” is made up of large state and teachers' funds, as well as a selection of large local funds. It is maintained by the National Association of State Retirement Administrators. Estimates for fiscal year 2011 funding levels by the Center for Retirement Research at Boston College project a small further decline in that year to 75 percent. It uses the same plans in its calculations as the “Public Fund Survey,” with the addition of the University of California Retirement System.

Key Findings: Retiree Health Care

 [i] The total bill for retiree health care benefits was larger than for pension benefits in only two of the 61 cities, as of fiscal year 2009. Boston and Bridgeport each faced more than a dollar in retiree health liabilities for every dollar in pension liabilities. (The unfunded portion of their retiree health liabilities also was larger than it was for pensions.) The reason is that both cities are responsible for retired teachers' health benefits but not their pension benefits, which are covered by the state. See “Bridgeport Comprehensive Annual Financial Report,” (June 30, 2011), 55, For Boston, see “Official Statement of the City of Boston Relating to Bond Offerings,” (2011), A33–A34,

 [ii] The following cities had larger unfunded liabilities for retiree health care than for pensions: Austin; Baltimore; Boston; Bridgeport; Charlotte; Columbia; Detroit; Fort Worth; Honolulu; Houston; Indianapolis; Jersey City; Memphis; Milwaukee; Nashville; New York City; Oklahoma City; Providence; San Antonio; San Francisco; San Jose; and Washington, D.C.

Lessons From the Recession

 [i] The annual recommended contribution (ARC) is the amount of money actuaries deem necessary to fund the benefits earned by active employees in any given year (dubbed the “normal cost”) in addition to money to pay down any unfunded liabilities. (The size of the payment also depends on a variety of other factors, such as benefits offered, contributions expected from employees, and the amount of time the city has chosen to achieve full funding.) Accounting standards currently establish the time for an unfunded liability to be paid down as no greater than 30 years, but some entities choose a different period. While some cities must make the total annual contribution because of law or pension system requirements, for others, the contribution often is not a required amount but a suggested amount, leaving cities free to underfund their benefits, thus raising the annual payment in subsequent years. To calculate the ARC, actuaries rely on assumptions for a number of future variables—the investment rate of return, inflation, and retiree life span, for example.

 [ii] According to Pew calculations, the funding level of the 35 cities that consistently paid at least 90 percent of their annual recommended contribution declined four percentage points between 2007 and 2009. The funding level for the remaining cities fell nine percentage points. 

Media Contact

Ken Willis

Officer, Communications