Saturday’s Wall Street Journal worries about how to “fix” 401(k) plans. The article discusses two distinct, widely disparate problems. One is that financial markets go up and down, so that someone who puts money into a 401(k) account runs the risk that his savings will decline in value. That is, of course, the point that a sophisticated grasshopper would have made to the ant. It’s an argument for making retirement income the responsibility of each generation’s offspring, with private thrift no more than a way to acquire luxuries. (In the Jiminy Cricket version of the fable, the grasshopper sings, “The world owes me a living”, demonstrating that he understands the implications of his own philosophy.)
The second problem (one that ought not to trouble those who regard the first as intractable, but who says pundits have to be consistent?) is that many individuals either can’t or don’t accumulate much through the 401(k) route. They may not be employed, or their employers may not sponsor plans, or they may not be attracted by the 401(k) tax incentives, or they may dissipate their accounts for short-term needs before retirement.
Evaluating the success or failure of section 401(k) (along with its various cousins: IRA’s, 403(b)’s, 457(b)’s, SIMPLE’s and SARSEP’s – how they differ is part of my professional life but needn’t interest you) should begin by taking an accurate view of their role in the provision of retirement income. While it isn’t literally inaccurate to call 401(k)’s “the backbone of the nation’s private retirement savings system”, for most workers they are less important than Social Security, and that importance declines with income. For low earners, Social Security replaces a majority of preretirement income. For example, a worker whose wages are consistently about half the national average can retire with a Social Security pension nominally equal to around 55 percent of his prior income; its actual value is greater, because it is (for a taxpayer with little other income) exempt from income and FICA taxes. Not much support from other sources is needed to boost these folks’ tax-adjusted income replacement ratios to the roughly 80 percent level that financial advisors typically recommend. A modest employer-provided pension or a relatively small nest egg will fill the gap.
Supplementing Social Security grows more essential as one’s income rises. While the low earner starts retirement over two-thirds of the way to the 80 percent threshold, somebody who has always had earnings equal to the FICA taxable wage base ($102,000 in 2008) would retire this year with a Social Security benefit of only about $27,000 a year. If he is to live at all comfortably in retirement, he will have to find twice that amount elsewhere.
Retirement income is, then, a middle class (and even upper class) issue. That tax incentives for saving are skewed toward the more affluent should occasion no shock; that is where incentives are needed. Looked at another way, the government provides well (wisely is a separate question) for the retirement of the bottom half of wage scale. No one expects (or should want) it to do the same for others. On the other hand, it ought not to tax away their efforts to help themselves.
Nonetheless, the remedies suggested by the experts whom the WSJ’s reporter consulted – an expanded savers’ tax credit, automatic enrollment in 401(k) plans, automatic employer contributions to IRA’s, etc. – will do the most where the need is least.
The Bush Administration – remember them? – tried to tackle this problem six years ago with a proposal for a three-tiered set of tax-favored savings accounts. One tier would have been a mildly modified version of the present 401(k) plan (relabeled “employer retirement savings accounts”). The others were called “lifetime savings accounts” and “retirement savings accounts”. LSA’s and RSA’s would not have been linked to employment. Anyone could contribute up to a maximum of one-half the 401(k) limit to each. Updated to reflect the 2009 limitations for 401(k) plans, the maximums would be $8,250 for an LSA, $8,250 for an RSA and $16,500 for an ERSA ($22,000 at age 50 and older).
The difference between LSA’s and RSA’s is that withdrawals from the former would have been permitted at any time for any reason, while RSA withdrawals before age 58, death or disability would have been penalized. The theory was that the existing penalties for premature withdrawals discourage contributions. Under the proposal, even the least provident would have a strong incentive to hold all but their shortest term savings in LSA’s, where they would earn tax-free income.
These concepts deserve revival. My own notion would be somewhat more expansive: Decouple 401(k) plans and similar vehicles completely from an individual’s employer, replacing them with LSA’s and RSA’s. I would set the LSA contribution limit at $10,000 a year, the RSA limit at $40,000. Employers could continue to make nonelective contributions to profit sharing plans (and, of course, to sponsor defined benefit plans), while employees’ elective contributions would move to these new vehicles. To overcome inertia, five percent of employees’ wages would be remitted automatically to LSA’s (from which they could be immediately withdrawn, so no one would be compelled to save). To minimize the short-term impact on government revenues, taxation would follow the Roth pattern of nondeductible contributions and tax-free withdrawals. The present value of the taxes is the same, whether they are collected on the front or the back end.
Many liberals have never liked section 401(k) and its cousins. In line with the maxim of never letting a crisis go to waste, they may see the budding recession as an opportunity to cast these disagreeable vehicles out in favor of something like Teresa Ghilarducci’s scheme for converting 401(k) plans into forced loans to the government.
Unhappily, no one has yet figured out how to remove risk from investing without also eliminating reward. I’m surprised, and mildly appalled, that Alicia Munnell, who certainly knows better, can declare, “As a major source of retirement income, [the 401(k) system) has shown itself unreliable – a point the financial crisis has driven home”. Saving for retirement is properly a project for one’s full working lifetime, a period of 35 years or longer. Those savings are then typically drawn down over 25 or 30 years. Over intervals of those lengths, bull and bear markets are of trifling importance. The great danger is long-term stagnation brought on by foolish government policies, and that is a danger that no retirement system, except for Anthony Trollope’s Fixed Period, can overcome.