Reforming the Defined-Benefit Pension System

In some respects the defined-benefit (DB) pension system in the United States is remarkably successful. In 2001 (the latest year for which data from Form 5500 filings are available), roughly $130 billion was paid in benefits to 11 million recipients. DB plans in the aggregate held an estimated $1.8 trillion in assets as security against future benefit payments, and in so doing made a large pool of capital available for efficient deployment through financial markets.¹ Over the thirty-plus years since the founding of the Pension Benefit Guaranty Corporation (PBGC), the government corporation that insures private sector DB plans, more than 168,000 plans have been terminated, but only in about 3,500 of those cases has the PBGC had to step in as trustee; in the other roughly 165,000 cases, plans had sufficient assets on hand to meet accrued obligations.² Even the often-cited $23 billion negative net position on the PBGC's balance sheet must be viewed in context, as it represents only a small fraction of total benefits paid since the founding of the insurance program in 1974.

In many respects, however, the system is not working as well as it should. One problem is that the system exposes some workers-those whose benefits exceed the maximum amount guaranteed by the government insurer-to a form of financial risk that they likely are ill positioned to bear and that they may not fully understand. In part the unsuitability of this risk reflects the fact that pensions loom large relative to many workers' financial portfolios. It also reflects the fact that a risky pension generally exposes workers to own-firm risk. (The own-firm nature of the risk inherent in a less-than-fully-secure DB plan arises because one of the preconditions of plan termination is that the sponsor must be in bankruptcy proceedings.) For some workers the downside of this risk has been realized: they have suffered severe blows to their personal financial situation when their employers have terminated their DB plans with assets grossly insufficient to pay accrued benefits, and the PBGC covered only a portion of their loss.

A second problem with the current system is that mispriced PBGC insurance coverage encourages excessive risk taking on the part of both workers and firms. Julia Coronado and Nellie Liang find that "even after controlling for cash availability . . . firms closer to bankruptcy have funded their plans much less generously."³ Anecdotally, other observers have cited cases in which firms in shaky financial condition have granted benefit increases in lieu of wage or salary increases, on the theory that if the fortunes of the firm improved, the firm would be able to meet the newly granted pension obligations, and if not, the government insurance program would step in to fill the gap.

A third problem is that investors seem to have considerable difficulty processing DB-related information accurately. For example, Coronado and Steven Sharpe document that investors seem to be much more sensitive to the smoothed information about pension items currently provided in the body of firms' financial statements than they are to the market-value-based estimates provided in the footnotes.⁴ Similarly, Francesco Franzoni and José Marín show that firms with severely underfunded pension plans-a publicly observable condition-tend to underperform the stock market.⁵ This misvaluation is a matter of macroeconomic concern because it raises the possibility that capital is being misallocated across firms.

A fourth problem with the current system is that it puts taxpayers at risk of providing substantial yet opaque subsidies to the sponsors and participants of DB plans, beyond the $23 billion already reflected on the balance sheet of the federal insurer. In particular, the Congressional Budget Office (CBO) estimates that the insurance provided by the PBGC will be under-priced by $63 billion over the next ten years if current law is left in place.⁶ This commitment of resources would come just as other demands on the federal government associated with the retirement of the baby-boom generation are mounting.