DRAFT--PRELIMINARY AND INCOMPLETE

I have been reading a nice article by Mary Williams Walsh about the withdrawal of corporate America from providing defined-benefit pensions and thinking about the implications for the Social Security debate:

More Companies Ending Promises for Retirement - New York Times: By MARY WILLIAMS WALSH: The death knell for the traditional company pension has been tolling for some time now. Companies in ailing industries like steel, airlines and auto parts have thrown themselves into bankruptcy and turned over their ruined pension plans to the federal government. Now, with the recent announcements of pension freezes by some of the cream of corporate America - Verizon, Lockheed Martin, Motorola and, just last week, I.B.M. - the bell is tolling even louder. Even strong, stable companies with the means to operate a pension plan are facing longer worker lifespans, looming regulatory and accounting changes and, most important, heightened global competition. Some are deciding they either cannot, or will not, keep making the decades-long promises that a pension plan involves.

I.B.M. was once a standard-bearer for corporate America's compact with its workers, paying for medical expenses, country clubs and lavish Christmas parties for the children. It also rewarded long-serving employees with a guaranteed monthly stipend from retirement until death. Most of those perks have long since been scaled back at I.B.M. and elsewhere, but the pension freeze is the latest sign that today's workers are, to a much greater extent, on their own. Companies now emphasize 401(k) plans, which leave workers responsible for ensuring that they have adequate funds for retirement and expose them to the vagaries of the financial markets.

One of the strangest--that is, the most inverted and self-contradictory--points made in the debate about social insurance roiling the world economy's industrial core is that claim that goverment-funded defined-benefit pension programs (like America's Social Security system) are outmoded: that they were fine for the industrial-economy age of the Great Depression and the post-World War II generation, but that they are definitely not appropriate for today's high-tech networked post-industrial economy. Just as corporations today are much happier contributing to workers' pensions in the form of contributions to workers' private accounts, so governments today should offer (or require) contributions to privately-owned accounts whose values fluctuate with the market rather than setting up a defined-benefit scheme that guarantees a fixed real sum of resources available upon retirement.

This is strange because it gets the economics of the situation exactly backward. When there are lots of companies offering workers long-term defined-benefit retirement pensions, there are fewer great advantage to the government setting up a parallel defined-benefit scheme and requiring workers to participate in it. Workers who greatly value a defined-benefit pension can select to work for firms that offer such pensions. The major benefits to having the government require that workers also participate in its own Social Security system accrue to those workers who really ought to highly value a defined-benefit pension but have not been able to figure out what their true preferences are, and to those relatively poor workers without the bargaining power to induce bosses to offer the pensions they really want, and need.

But today there aren't a lot of companies willing to offer long-term defined-benefit pension schemes. One reason is that companies are now much more aware of their own long-run fragility than they were in the post-World War II decades. Not even America's IBM wants to take the risk, any more, of offering defined-benefit schemes. The risk used to be less than three offsetting benefits to a company that offered a defined-benefit pension: the fact that leaving the company usually meant you cashed in your pension at a discount increased worker loyalty, the firm's greater size gave it a greater risk bearing capacity, and complaisant accountants making optimistic assumptions about returns on pension reserves brightened the financial picture companies could report to Wall Street. Today the risks are seen to be much greater. And the benefits are seen to be less. Hence a smaller and smaller slice of American employers are offering anything like defined-benefit pensions.

This fall-off in private defined-benefit pensions all across the rich core of the world economy is a bad thing, for everything we can read off of the configuration of asset prices suggests that defined-benefit pensions are things that young and middle-aged workers value extremely highly. Historically, the gap between expected returns on low-risk assets like government or investment-grade bonds and high-risk assets like stocks and real estate has been very high. To some degree this may be because (as economists like Robert Barro and mathematicians like Benoit Mandelbrot argue) high-risk investments are in reality much more risky than the theories and math of standard finance techniques suggest. To some degree this is--in my opinion, at least--because the memory of years like 1930 and 2000 when stocks did very, very badly indeed occupies too large a place in investors' brains. To some degree this is because workers and other asset holders do indeed place a very high value on safety, security, and predictability. Thus offering a defined-benefit pension plan to workers is an extremely valuable thing to do.

But in today's world, only national governments are large enough to be able to do so with any assurance that the pension assets will actually be there when workers retire.

I am enough of a social democrat to believe that if there is an economic service or benefit that citizens value extremely highly and that only the government can provide, then the government should provide it. We economists know that there are many drawbacks to expanding government beyond its basic roles of defense, public safety, justice, true public goods, and providing citizens with incentives to counterbalance the effects of true market failures. If the private market has the flexibility of two hands, the government bureaucracy has at best two thumbs. But the collection of payroll taxes from tens of millions of workers and the writing of tens of millions of pension checks is the kind of routinized, semi-automatic task that the government can do well. And it is much more important and valuable that it do it in our future post-industrial network-age society than it was in the past.