Funding Public Pensions

Common "Solutions" That Seldom Solve Anything

Common "Solutions" That Seldom Solve Anything

The accounting rules for pensions as described by GASB are not merely poor indicators of a system’s financial health and stability, they are an equally poor guide to remedial action. What appear, under these rules, to be sensible strategies for improving a pension’s funding are, when judged by practical experience, highly risky and speculative ventures that routinely create catastrophe where none was necessary.

Obviously any plan with a funding shortfall can simply modify the contributions to the plan or the benefits paid from it, and these solutions are legion: raise the retirement age, lower the benefit, raise the employer contribution, raise the employee contribution. Some plans have opted to create tiers of employees, to economize by providing lower benefits for newer hires. These are all part of the standard austerity playbook, and adequately covered in many texts on the subject.54

But a few other options exist and are widely used, often to the detriment of the policy makers who choose them. These include (1) pension obligation bonds, (2) closing a pension plan, and (3) shifting employees to a defined contribution plan. Each incurs risks at least partly obscured by the accounting rules. A partially funded pension plan that uses one of these strategies takes on this hidden risk and therefore may face a higher risk of funding shortfalls in the future than a similar plan that forgoes them. 

Pension obligation bonds: Investing on the margin

The liability of a pension is a debt owed to the future retired members of a pension plan. The fund assets are the wherewithal currently on hand to pay that debt, so the unfunded portion of that liability can be thought of as an unpaid debt compounding at an interest rate equal to the discount rate used to forecast liabilities.

According to this analogy, the sooner it is paid off, the better. If it is possible for a government to borrow at a lower rate than that discount rate—and for most governments it is—an obvious strategy is to borrow at the low rate in order to pay back the debt compounding at the higher rate. This is a pension obligation bond (POB). 

However, another way to look at it is that a government that issues a POB is borrowing on the margin, and betting the returns achieved will be better than the interest to be paid. This is highrisk investing. Oakland, California, issued the first such bond, in the 1980s, and Congress quickly acted to say that such bonds are not tax-exempt. Pension bonds are thus market rate bonds. Since the governments that issue them tend to be the troubled governments, whose bond ratings are likely not AAA, the margin between the interest rate paid and the pension fund discount rate tends to be small.55

Imagine a pension fund with $2 billion in assets, with an unfunded liability of a billion more, and thus a funding ratio of 67 percent. Assume all goes very well, and the government borrows that extra billion at 5 percent over 30 years, while managing to invest it all at an average of 7.5 percent over the whole term. This sounds good, but another way to look at this is that the fund actually earns 7.5 percent on two-thirds of the fund and only the difference between that number and the POB interest rate, or 2.5 percent, on one-third of the fund. This creates an effective rate of return of 5.83 percent, much less than the assumed rate. In other words, the cost to the government of this pension plan is the same as a fully-funded plan that is not achieving its investment targets. This is hardly a knock-out argument in favor of such a bond.

Furthermore, the bond capital must be repaid, not just the interest. If the POB payments capture the bulk of what might have amortized the unfunded liability, and the government and employee contributions just barely cover the normal costs of plan members, the plan may not be in appreciably better shape at the end of the POB term than it was at the outset. That is, after a 30-year POB has been paid back, the funding ratio for our example fund might not improve at all—even if everything goes as well as can reasonably be expected. This will seem especially harsh since, for the previous 30 years, policy makers will have regarded the plan as fully-funded. The temptation simply to roll over the debt by issuing another such bond will be very strong. 

Of course it is very seldom that everything goes as well as can be expected. In reality, financial market returns are volatile, and POB margins are small, so the success of such a bond depends heavily on the market timing. In a survey of several thousand POBs, researchers with the Center for State and Government Excellence found that, as of 2014, of the POBs issued in the previous 20 years, most were only barely in positive territory, and a few had lost quite a lot of money. Timing and luck appear to be the determinants, awkward components out of which to build sensible policy.56The Government Finance Officers of America is fairly blunt in their assessment (part of their collection of “best practice” documents): “The GFOA recommends that state and local governments do not issue POBs…”57

Even more unfortunately, the fixed borrowing costs of any bond market transaction encourage governments planning to take such bets to take large ones. Detroit, Michigan, sold $1.44 billion in POBs in 2005—and lost $2.8 billion on the deal.58 Stockton, California, sold $125 million in POBs in 2007, and had to go into bankruptcy when the gamble failed. The fund was left in worse shape than before the bond was issued, and the city owed the POB capital to its bondholders.59

Closing a plan: Riskier than you might think

Many political leaders, faced with an apparently colossal pension debt have concluded that closing their government’s pension system is the only feasible alternative. The obvious problem with such a strategy is that it is not an answer to the question of funding the debt. That is, closing the plan because the debt is too large does not excuse a government from paying that debt, even if it does keep the debt from growing further. 

Closing a pension plan has important risks that may not rise to the awareness of a policy maker looking at the raw numbers. These risks are again masked by the accounting rules. The first is simply that one of the real strengths of a pension system, essentially unacknowledged by the GASB rules, is the flow of money into the system from its members and their employers. Depending on the financial market conditions, this flow is usually a comparable size to the investment returns or larger. The consistency of this contribution makes it at least as important in providing security to the system as investment earnings, if not more so. Removing this source of strength leaves a pension plan to rely on the financial markets alone, not a well-known source of security. 

Beyond these concerns, there are investment constraints imposed by closing a plan. One of the advantages of managing investments in perpetuity is that managers are permitted to take the long view in everything. Earning the necessary 7.5 percent may be feasible when all investments can be long ones. A system without that long horizon has a serious disadvantage when it comes to the array of investments open to it. As a pension plan winds down, more and more of its fund must be kept in relatively liquid short-term investments. These seldom earn the higher returns of the longer-term assets. Therefore, the greater the proportion of the fund that must be kept in short-term investments, the less feasible it is to achieve the typical investment return targets. A 2011 analysis done by the California Public Employees’ Retirement System (CalPERS) showed that closing that system would give up $150–200 billion in investment returns over the course of winding the system down, about half the size of the overall fund.60

Sometimes, closing a plan is done with a POB. This is subject to all the risks of a POB, with the added risk of ending the employee payments into the fund. If a POB is a tightrope act, only successful when interest rate and market conditions are just right, closing a plan with a POB is a tightrope act without the net. The city of Woonsocket, Rhode Island, issued a $90 million POB to close its police and fire employee pension plan in 2002. A dozen years later, as reported in their 2014 financial report, after some poor investment years, the plan has assets of only $46.4 million, still has an unfunded liability of $42.2 million, only a single employee still paying into the system ($5,000 each year, compared with $8 million in expenses), and the city has a $79.4 million left of the bond debt to repay.61

Defined contribution pension plans: Not really a pension, or a solution

Another common strategy to addressing an unfunded pension liability is to transition to a DC plan, or a hybrid pension plan with DB and DC components to it. The claim here is that a government can shed some of its risk by moving employees to that kind of pension. Certainly the inflation risk and the investment risk are with the employee under a DC plan. But converting from DB to DC is approximately the same thing as partially closing a plan, so the risks of closing a plan are not avoided, merely reduced. Beyond those risks, there are others in this approach. 

To begin with, the administrative costs tend to be higher for a DC plan than for a DB plan. In the CalPERS analysis cited above, the authors point out that fees for their DB plan amount to about a quarter-percent per year, while a typical DC plan charges between one and two percent per year.62 Thus the investment returns must be that much higher to achieve the same standard of living for the retiree. 

Large plans, the size of a state or big city, can usually negotiate substantially lower fees for a DC plan,63 sometimes only a few basis points higher than a comparable DB plan if the options for the employees are few and simple, but other investment constraints still offer cause for concern. Because the investments are being made on behalf of an individual with a finite lifetime rather than a perpetual group, more must be saved, and the returns will be lower. 

Isolated from a group, an individual employee must save for his or her maximum expected life expectancy, not merely for the average, as with a DB plan. Paradoxically, this not only increases the risk of underfunding a retirement, but also increases the risk of overfunding as well. That is, it is harder to reach the appropriate target, so fewer people will succeed, and the appropriate target is too high, so only the luckiest of the successful—the ones who both save enough and live their maximum life expectancy—will not be wasting money. Everyone else who has successfully saved enough for their maximum life expectancy will have had earnings they were unable to enjoy while alive. Their heirs may not consider this a problem, but from the perspective of efficiency, it is a very real waste of resources. And these are only the lucky few; there will be substantial hardship among the many who have not saved enough. 

The CalPERS analysis points out the other problem with investing for an individual, that whereas a traditional pension plan can be invested for a very long horizon, a DC plan must adjust its investments for the ages of its members. An older member needs a more conservative investment mix than a younger member—and a more liquid mix once the member retires—and this reduces the potential rate of return, increasing the risk that an individual’s savings will be inadequate to finance a comfortable retirement. 

In addition, many pension plans incorporate some form of disability or survivor’s benefits. A government that must honor that commitment to its pension plan members will see an additional cost to provide the same benefit to its DC members. 

Claims that money can be saved by converting from a pension plan to a DC savings plan usually rest on the argument that the costs of running a pension plan are underestimated, rather than on a denial of these transition costs.64 Other common arguments are that the assumed rate of return for a pension plan is unrealistic, or that the lowered investment returns from the shortened time horizon are inconsequential, if not negligible.65 As we have seen, the existing accounting rules tend to exaggerate the costs of a pension plan. It is that very exaggeration on which these arguments rely to make their point about the relative savings of closing a pension plan, or the advantage of DC plans. 

Another argument is frequently made, that private industry has “moved away” from the un-sustainable costs of pension plans. This is a tendentious reading of recent history. Private defined-benefit pension plans across the country have been closed not because they were unsustainable, but because the executives of the corporations sponsoring them wanted to use that capital for different purposes and existing law provided a way for them to do so.66 The 1980s saw a blossoming of the private equity industry, devoted to buying companies and repurposing their financial assets to the benefit of the new owners. The decade also saw repeated raids on the pension funds of the newly purchased companies, raids that continue to this day. More recently, congressional changes to the law governing pensions and the Pension Benefit Guaranty Corporation provided incentives for corporate managers to end these plans and turn the responsibility for their retirees over to the government, a managerial moral hazard. But even as of 2000, when the most recent wave of disassembly began, most remaining corporate pension funds were fully-funded or even overfunded. The stock market losses of 2000–2001 cut into that margin, but not significantly. It was the desire to repurpose that capital, combined with the federal guarantee and new rulings about bankruptcy, that made ending these plans desirable, and thus inevitable.67

A report from the National Institute on Retirement Security listed several reasons for the decline of DB pension plans, including decreasing union presence in the work force, the legal and regulatory environment, and changes in corporate priorities, and added: “It is interesting to note that each of these reasons has little to do with the underlying economics of maintaining DB pension plans.”68

In truth, given that a DC plan does not envision a level of retirement income at all, it is perhaps arguably not even appropriate to label it a “pension” plan at all. In fact, the phrase was not used until the 1970s. Until then, it would have been called a savings plan, or a stock purchase plan, if it had a name at all. In many ways, the rebranding of a simple savings plan as a “pension” plan is a triumph of marketing in service of a massive reappropriation of pension plan assets. 

  • 54. For example: Peng, State and Local Pension Fund Management; Winklevoss, Pension Mathematics with Numerical Illustrations.
  • 55. Alicia H. Munnell, Jean-Pierre Aubry, and Mark Cafarelli. An Update on Pension Obligation Bonds. Tech. rep. 6/20/15. Washington, DC: Center for State and Local Government Excellence, July 2014. URL:
  • 56. Ibid.
  • 57. Government Finance Officers of America, Advisory, Pension Obligation Bonds, 6/26/16, The advisory goes on to list five very good reasons to avoid POBs: The invested proceeds might fail to earn more than the interest rate. POBs commonly have a complex structure with derivatives embedded in them that introduce counterparty risk and credit risk. POBs are taxable debt, which usually counts against a government’s overall debt burden for the ratings agencies. The principal (re)payments may be missing, or structured over a longer term than the amortization period. POBs are a red flag to ratings agencies, who generally take them as a sign of other trouble, potentially undisclosed.
  • 58. Though this may have been mostly to do with the city’s ill-fated attempts to hedge the floating interest rate of the POBs it issued by swapping with banks for fixed-rate debt payments. See Farmer, “Detroit’s Pension Is Actually Well-Funded, So What’s All the Fuss?.”
  • 59. Walsh, “How Plan to Help City Pay Pensions Backfired.”
  • 60. CalPERS. The Impact of Closing the Defined Benefit Plan at CalPERS. 6/21/15. California Public Employees’ Retirement System (CalPERS). Mar. 2011. URL: http://
  • 61. The plan has almost 60% of its assets in cash and fixed-income investments, so it takes a remarkable investment year for its assets to earn the 7.5% it needs to match its assumptions. In 2014, however, its assets did earn 8.28% gross (not counting investment fees). See the 2014 Woonsocket annual financial report at 2014.pdf (6/25/16), pp.58-65. The annual contributions of the single remaining employee were noted in the 2013 report ( Woonsocket_2013.pdf). See also Sandy Seoane, “Despite concessions, Woonsocket’s pension is forecast to run dry.” Valley Breeze, December 2, 2015. 6/26/16, URL: woonsocket-north-smithfield/despite-concessions-woonsocket-s-pension-forecast-run-dry
  • 62. CalPERS, The Impact of Closing the Defined Benefit Plan at CalPERS.
  • 63. Deloitte Consulting. Inside the Structure of Defined Contribution/401(k) Plan Fees, 2013: A study assessing the mechanics of the “all-in” fee. Tech. rep. 9/29/15. Washington, DC: Investment Company Institute, Aug. 2014. URL: dc_401k_fee_study.pdf.
  • 64. See, for example, Andrew G. Biggs, Josh McGee, and Michael Podgursky. Transition cost not a bar to pension reform. 6/21/15. Jan. 2014. URL:
  • 65. Lance Christensen, Truong Bui, and Leonard Gilroy. Addressing Common Objections to Shifting from DefinedBenefit Pensions to Defined-Contribution Retirement Plans. 6/21/15. June 2014. URL:
  • 66. Fran Hawthorne. Pension Dumping: The Reasons, The Wreckage, The Stakes for Wall Street. New York: Bloomberg, 2008.
  • 67. Ellen Schultz. Retirement Heist: How Companies Plunder and Profit from the Nest Eggs of American Workers. Portfolio/Penguin, 2011.
  • 68. Ilana Boivie. Who Killed the Private Sector DB Plan? Issue Brief. 6/12/15. Washington, DC: National Institute on Retirement Security, Mar. 2011. URL: