April 2006

The PBGC and Structural Reform: With a $23 Billion Deficit, Where Do We Go from Here?

The 2005 Pension Benefit Guaranty Corporation (PBGC) annual report shows that its liabilities are $23 billion greater than its assets. With bad luck, experts believe that this could grow to $100 billion and result in a taxpayer bailout on the order of the savings and loan collapse. The PBGC problem results not from unfortunate accidents, but from structural issues reflecting global economic changes, demographic shifts, the nature of open-ended government “insurance” of financial obligations, and numerous design problems in the structure of the PBGC itself. Panelists at an April 5 AEI conference discussed what can be done to protect taxpayers, and whether defined-benefit plans tied to a government guaranty have a future.

James K. Glassman
AEI

There is currently a story by Steven Ellis in the Australian newspaper titled “Pension Liability Crisis Looming.” If GM ceased operations tomorrow, it would leave unfunded pension and health liabilities of about $69 billion--twice the firm’s cash reserves after the GMAC sale, and four times its shareholders’ currently recorded equity. The U.S. government would end up picking up the bulk of these obligations through the PBGC, which is already in a $23 billion hole, after having to assume much of the pension liabilities of U.S. steel and airline industries. Such liabilities have increased as a string of private firms have walked away from their under-funded defined-benefit pension schemes.

The PCBG was created by the Employee Retirement Income Security Act (ERISA) in 1974, and currently protects the pensions of 44 million Americans in 30,000 private single-employer and multiple-employer defined-benefit pension plans. The PBGC receives no funds from general tax revenues, and operations are financed by insurance premiums set by Congress and paid by sponsors of defined-benefit plans and recoveries from companies formerly responsible for those plans.

To cover operations for the next seventy-five years, a bailout of $100 billion would be required. The Savings and Loans (S&L) crisis comes to mind. Like the S&L crisis, the PBCG problem is the result of structural issues and is loaded with moral hazard--providing the wrong incentives for the companies that maintain defined-benefit pensions. It must also face changes in the global economy and demographics at home. Companies like IBM are increasingly moving towards defined-contribution plans. Will the only plans left for the PBGC to cover be the weakest? There is a Congressional conference committee working on some PBGC reform legislation, with a deadline in May. Will this make things better or worse?

Bradley D. Belt
Pension Benefit Guaranty Corporation

Whenever you have public sector backing of private sector risk-taking, problems may arise, and if you want to avoid moral hazard, you must make sure that there are checks and balances to avoid imprudent risk-taking or cost-shifting.

The core mission of the PBGC is to provide financial guarantee insurance to pension plans. If you compare and contrast this to private sector insurance, some of these infirmities show up pretty quickly. The PBCG cannot set underwriting standards, as a private sector insurer would do.  We are not able to adjust premium levels to compensate for actual or projected claims. We are not able to set out premiums in a risk-based fashion, to protect against the riskiest behavior and to encourage appropriate behavior. We have to provide insurance coverage no matter how uninsurable the risk is, regardless of whether a contributor makes payments into the system or makes contributions to its pension plans. Most remarkably, the insured can cause the insurable event and still collect on the insurance policy. One does not wonder why the PBCG has a $23 billion deficit with that kind of business model.

With United Airlines post-9/11, when the company and industry were in some financial difficulty and pension plans were in many multiple billions of dollars under-funded, the workers still negotiated $800 billion in benefit improvements as deferred compensation, guaranteed by the PBGC. During the 1980s and 1990s, the robust economy obscured from view these structural flaws, equity markets were sailing along, and everything was fairly smooth. But we have belatedly come to recognize potential adverse consequences of the manifestations of these structural flaws.

There are 44 million Americans covered by defined-benefit pension plans insured by the PBGC, and the PBGC is now responsible for providing benefits to 1.3 million participants of plans that have now terminated.

We have seen that the pension insurance program has a major impact on American companies and entire industry sectors. Pension obligations have become so large and under-funded that there is a substantial amount of cash flow going toward meeting these obligations that would otherwise have gone to funding productive activity elsewhere in the enterprise.

Some companies have taken advantage of the interaction between the bankruptcy code and the employment security income to offload their pension obligations to the pension insurance corporation, go into bankruptcy, and reemerge as a business shorn of these legacy labor costs. The problem is not only that they have shifted these costs to responsible third parties that may have to pay higher premiums, but they have gained a competitive leg-up over their competitors, who, in turn, are pressured to do something about their own legacy costs. ERISA was not designed to provide a competitive advantage to certain companies that wish to unload their legacy costs, but that is how it has been used more recently.

These issues affect investors. Equity has often been wiped out because of difficulties that companies have encountered with their pensions. If equity investors are not aware of these issues, then how are they going to effectively allocate capital and make pricing decisions with regard to lending to particular enterprises? There is a quality of earnings problem. Earnings have often represented returns on pension assets rather than skill in building productive activity. To be able to discount future cash flows when valuing an enterprise, you also have to be able to figure out what those future cash flows are going to be. This information is embedded in ERISA rules but is not showing up in financial accounting statements. If you do not know whether a company has just been skating along and avoiding deficit-reduction contribution rules (although these will eventually catch up with them and there may be a falling-off-the-cliff effect), it is very difficult to determine what the cash flow of the enterprise is going to look like down the road, and therefore very difficult to determine the appropriate enterprise value. So, there is a real risk of misallocation of capital.

Beyond individual investment decisions, we are talking about an aggregate of about $4 trillion in defined-benefit plans in private and public sectors, which are allocated and invested, and these have real implications for future interest rates and equity prices. In the UK, more recently, we have seen some shift out of equities into fixed income and alternative asset classes.

We must recognize these infirmities. It is the status quo that has led us to a continued erosion out of the defined-benefit system and this $23 billion fiscal gap at PBGC, hundreds of billions of dollars of under-funding of pensions, and the offloading of pensions by companies like United Airlines with a $10 billion funding gap and lost benefits. These outcomes are unacceptable.

The administration has proposed strengthening funding rules. We would not need to be having this panel today if United’s funding gap was much reduced. We need to have a rational risk-based premium structure that puts the PBGC on a long-term solvency path, rather than large and growing deficits. We also need to have more transparency so that stakeholders--such as workers or retirees, investors and regulators--have regular and timely information to make more informed decisions.

Douglas J. Elliott
Center on Federal Financial Institutions (COFFI)

The PBGC owes $23 billion more than the value of the investments that it holds, and this number actually rises by the month. When firms make payments according to a defined-benefit pension plan, the PBGC stands as a guarantor of that. But that guarantee only comes into play if the company goes broke and does not have enough money put aside in the pension plan to cover the promises. The PBGC is the third line of support, but when you are supporting $2 trillion in obligations, it does not take a lot of bankruptcies (especially given the moral hazard issues) to produce a very large debt.

To work out the precise debt figure, you have to bring future liabilities into today’s dollars by using an interest discount rate. Whatever rate you use, you cannot plausibly argue that the PBGC has enough money now to pay for all the promises that it has to take over. It is therefore insolvent, even though it has $60 billion in investments.

At COFFI, we have a set of models that, on the basis of current law and future trends, predict that PBGC will run out of money in 2022. Unlike Social Security, where continuing payroll taxes can fund 75 percent of the promises, future PBGC premiums would only cover about 9 percent of future payments, and so funding would fall off a cliff. Politically, that will not happen, even though there is no obligation to the taxpayer to bail the system out. There is no way that Congress is going to let grandma and grandpa have their checks bounce.

There is a structural imbalance between the premiums and the risks. The most optimistic academic study suggests that premiums would have to be doubled in order to take care of PBGC’s promises. Although this may be true historically, it does not account for the fact that there are a number of losses that are likely to come in over the next few years. There is an analysis that suggests that if GM went bankrupt, there could be a claim on PBGC of $31 billion, as they stand today. But one of the bad things with the pension insurance system is that companies that have troubles with under-funding tend to have even more problems by the time that it actually becomes a claim on PBCG. So, it is not crazy to think that this $31 billion problem is actually a $50 billion one.

The best-case estimate is that you would have to inject $92 billion in today’s dollars in order to cover the existing problems, plus make sure that it could pay out over the next seventy-five years the promises that it will make in the next few years. The Congressional Budget Office’s more pessimistic models show that if you had to persuade a private insurer to take on these promises (including the current $23 billion hole), you would have to pay the insurer $142 billion to cover the next twenty years of claims.

The administration has proposed the three ways for structural reform that Bradley Belt outlined, and everyone will agree with the premise that these three goals are laudable goals. There is debate regarding the extent to which the actual bills that are being reconciled in conference meet these objectives.

There are two key goals that are possibly in irreconcilable conflict. We want to avoid a taxpayer bailout, but we want to encourage companies to voluntarily offer these kinds of pension plans. The only way to help PBGC’s finances is to shift costs and risks back to the employers. Employers are already uncertain that defined-benefit plans are worth the costs and risks that they already bear, and so anything to help PBGC’s finances may discourage firms from continuing to offer these plans.

Allan I. Mendelowitz
Federal Housing Finance Board

The history of past government bailouts has not been encouraging. When you use taxpayer money, nobody has responsibility for it, and no one has any incentive to pay attention to its use.

Looking at the Chrysler loan guarantee program (known in the press as the “Chrysler bailout”), this has very little in common with other such interventions. Substituting government involvement for the responsibilities of a private company inevitably shifts decision-making from bankruptcy courts to the political process. Bypassing these bankruptcy courts weakens the powerful incentives that we have in our economy for promoting good management and rationalizing the use of capital. However, if the government is going to make taxpayers liable for some of the costs of bankruptcy, it must be done right. I believe that the design and execution of the Chrysler program met this test and that its principles could form the basis of PBGC reform.

One of the justifications given for the Chrysler loan guarantee program was that it became clear that it would be cheaper for the government to keep the company in business through a well-designed bailout than to let it go bankrupt and have to take over its pension programs. This shows that public funds can be put to work in a way that is responsible, efficient, and effective.

For success, there has to be a clear understanding of the objectives of the program. I once authored a paper on the PBGC: “A Program in Search of a Policy.” If you are spending money, you need to know why and what you want to get for it, and any well-designed bailout must contribute to furthering the public purpose, rather than just flowing through to augment economic rents of shareholders.

With the Chrysler deal, every group that stood to gain from the survival of the company had to contribute to the financing of the turnaround. Indeed, two dollars were required from other sources for every dollar that Chrysler received. Chrysler shareholders were not allowed to free-ride on the bailout, and their holdings were watered down. To secure the loan guarantee, Chrysler was also required to be a viable business going forward. The guarantee was structured so that the federal government gained priority over every other creditor of the corporation, with first claims over all its assets. The Chrysler program did not become a treasure map for those private sector businesses trying to avoid the consequences of poor management.

The problems at PBGC result from structural changes in the U.S. economy, which have challenged basic assumptions behind traditional defined-benefit plans, government regulations, actuarial requirements, and incentives that have led to expensive future obligations. Within the institution, there are structural issues regarding a lack of actuarially based prices and perverse incentives that have created a growing financial hole.

Is the objective to reduce the financing gap or reform the perverse incentives? There are complex trade-offs involved in achieving these goals. Raising premiums may increase the cash flow in the short run, but may lead to defined-benefit plans being shut down or closed out.

None of the proposed plans look to the beneficiaries--the workers whose plans are protected. The Chrysler plan called on the beneficiaries to contribute the required resources. We should follow this principle by calling on beneficiaries, rather than operators, to contribute to funding the insurance of benefits.

What government does for pension beneficiaries is analogous to what it does for flood insurance. The federal government forces mortgage-holders to take out homeowners’ insurance to protect the value of the collateral in case of fire or something else that may destroy the structure. It is the buyer of the insurance, not the bank, who is required to buy the flood insurance (which the market cannot provide), along with their normal insurance. The beneficiaries of defined-benefit plans are the same of homeowners, in the sense that they receive an asset that has a substantial amount of risk associated with it. Working on this basis would allow you to close the financing gap for around $10 per month.

The average beneficiary of a PBGC plan is wealthier than the average taxpayer. You would therefore find poorer taxpayers supporting wealthier ones.

Alex J. Pollock
AEI

This giant financial problem goes back to an idea from March 1961, when Nat Weinberg went to Walter Ruther of United Auto Workers (UAW) and said something like this: “We could have the equivalent of deposit insurance for our unionized pensions. This way we would not have to negotiate with the companies to fund our pension plans. We could get the government to guarantee them, and we could use our chits to negotiate other benefits.” Later, in 1974, this was enacted.

This came out of the discovery that big established companies could fail, and when they failed, their pension plans would not have enough assets, leaving the pensioners without income. As others have pointed out, when a company sinks into financial trouble, pension funding is (quite understandably) a low priority for the management. If creditors are hounding them, management is not going to sink the cash that it can use into pensions. Moreover, the cost of pensions can become one of those factors dragging companies down, as we have seen with legacy costs. So, we must say that Mr. Weinberg’s idea was truly brilliant in terms of solving the UAW’s problem, but has given us a problem later.

Now the PBGC, which guarantees these pensions, is insolvent. When a government issues open-ended guarantees on someone else’s financial performance and managerial prowess, it is extremely risky. The Book of Proverbs says: “He who gives surety for a stranger will smart for it.” With the PBGC, promises have been made, and it is now time to finance them. In a sense, companies have borrowed money from employees in return for future promises to pay.

Congress is supposed to promote defined-benefit plans, with all the problems that we have talked about. It is not just meant to be a guarantor; it is supposed to be a cheerleader. It must take the substantial risk of these plans, while keeping premiums low--at a rate that it cannot control. These three goals are manifestly in conflict with each other and cannot all be done.

You can raise premiums on somebody (and I like Allan Mendelowitz’s idea that it should be the beneficiaries), you could force companies to put money into their plans faster, you could take tax money; or you could do all of the above. Out of the 143 million people employed in the United States, 22 million are covered by PBGC plans. Should the 85 percent who do not have these rich guaranteed pensions pay extra taxes to bail out the 15 percent who do? That does not seem like the greatest political calculus.

If you raise premiums, you make it less attractive for companies to sponsor defined-benefit pension plans. Strong companies very justifiably ask: “Why should we be paying all this money to bail out the weak companies that have made mistakes and been poorly run?” And the weak companies ask: “Why do you kick me when I’m down? You are making bankruptcy even more likely.” The more you do to solve the funding problem, the more you do to hasten the demise of defined-benefit plans.

Companies with defined-benefit plans have actually gone into two completely different businesses, operating their normal business and the business of offering and underwriting life annuities. The bigger their pension plan gets, the larger the annuity-writing business gets relative to making cars or flying airplanes or extruding plastic parts or whatever it is that they have their distinctive competence in. In 1970 when General Motors had a 50 percent market share, the pension and annuities business was a small fraction of their overall operations. Today its market share is around 25 percent, and the company has a huge annuity-writing business relative to their automobile operations.

There are fundamental risks associated with writing annuities, such as shifts in life expectancy and interest rates. Indeed, life expectancy has been underestimated, and interest rates have been overestimated. Additionally, retirement may be earlier than you expected, particularly if the company faces financial difficulty and has to downsize. If you allow some people to take lump sums out of the plan, they get 100 cents on the dollar, and everyone else is left with less. This can work the same way as a bank run.

Yet, even though this annuity business has risks, there is no explicit capital requirement and there is not much prudential regulation. Pension funds should be conceptualized the same way we view the writing of life annuities and have the same kinds of regulations regarding capital requirements and investment risk applied to them.

It is not clear what assets these defined-benefit plans have with respect to their obligations. They essentially borrow from their employees and invest the money into equities. This may be profitable in the long run, but there is no way of seeing whether there is sufficient security.

These have a large component of cross-corporate equity holdings. We criticized Japanese companies a lot after the stock market bubble for holding each others’ equity. This created a feedback loop, which caused companies to seem more profitable as the stock market went up, with an opposite multiplier as the market went down. But exactly the same thing happens with the American defined-benefit pension system.

So, what is our objective with the PBGC? Is it to promote defined-benefit plans as a political idea, or to have this system run on a financially sensible basis? James Glassman has written that “defined-benefit plans are headed for the dustbin of history.” I believe that our real mission is no longer to promote defined-benefit plans, but to help smooth the transition of the American corporate sector out of these plans, which create so many structural problems.

AEI research assistant Chris Pope prepared this summary.