401(k) Income Projections Are a Good Idea
Alicia H. Munnell is a columnist for MarketWatch and director of the Center for Retirement Research at Boston College.
Showing the connection between saving and income in retirement could improve saving decisions
The U.S. Department of Labor issued for comment an advanced notice of proposed rulemaking that would encourage the sponsors of 401(k) plans to project how much income a participant’s 401(k) balances would provide at retirement. Shifting the focus from stocks (401(k) balances) to flows (monthly income) will provide participants with a much better sense of the portion of required expenses that their 401(k) accumulations can cover once they stop working. The hope is also that by showing the connection between saving and income in retirement that people might make better saving decisions.
The most recent data show that the median household approaching retirement (age 55-64) with a 401(k) had $120,000 in their combined 401(k) and Individual Retirement Accounts (IRA). (IRA balances are included because much of the money in these accounts is rolled over from a 401(k).) To many, assets of $120,000 may sound like a fortune. But, depending on the drawdown strategy, this amount will provide only about $500 per month.
In its request, the Department of Labor asked for comment on whether income data should be provided and how best to present accrued benefits as an estimated stream of lifetime income. (Comments were due July 8, 2013.) Some might say that precise calculations are too difficult, so any information would be misleading. Such a dismissal would be a mistake, however. While any calculation involves a number of assumptions, a sensible estimate is better than nothing.
And the Department of Labor makes a stab at how such projections could be presented. It proposes two projections. The first is the income that participants’ current account balances would provide if they were now at the normal retirement age. The second is the income they would receive at retirement assuming their current balance would grow with future contributions and investment returns. The assumption is that contributions would increase by 3 percent per year and that the nominal return is 7 percent. The projected balance would then be deflated by 3 percent a year to get it back into today’s dollars. Eliminating inflation from the calculation is extremely important or else everyone would look like projected millionaires. Social Security’s online benefit statement calculations also present projected benefits in terms of today’s dollars.
Projections for Social Security benefits are much easier than for 401(k) balances once people arrive at retirement. Social Security has a benefit formula, and the provisions of that formula can be applied to the individual’s 35 highest years of average indexed monthly earnings. In contrast, the projections for 401(k) balances yield a lump sum, and assumptions are required to determine how much that lump sum can produce in terms of income.
The Labor Department appears to be suggesting that participants use their money to buy an actuarially fair annuity – that is an annuity priced for the average individual (as opposed to those whose parents died in their 90s) and without marketing and administrative costs. This assumption may provide a somewhat optimistic estimate for two reasons. The most straightforward is that people do not purchase annuities with their 401(k) balances. A whole sub-specialty has emerged in economics to explain people’s lack of interest in annuitization, even though, through pooling, annuities can provide a higher level of income than people can earn on their own. If 401(k) participants instead use the “4 percent rule” or some other rule-of-thumb to withdraw their balances, their monthly amounts will be considerably lower. Less importantly, any annuity that individuals can actually purchase in the private sector will provide less income than an actuarially fair annuity.
A recent study published by the Center reported on a field experiment that tested the effect of retirement income projections on saving decisions, involving 17,000 employees at the University of Minnesota. The study showed that providing individuals with retirement income projections, along with related material on retirement planning, modestly increased saving at very low cost.
Despite the complexities and the myriad of assumptions involved both in projecting balances at retirement and monthly income thereafter, reporting current and projected balances in terms of a lifetime stream of income is a useful step forward.