Tweet This
The Center for Retirement Research at Boston College released a nice paper by Steven Sass looking at whether the Social Security Statement improved Americans’ knowledge of the benefits they’ll receive once they retire. Since 1995, the Social Security Administration has sent working-age individuals a personalized projection of the benefits they can receive in retirement. (You can also access those estimates online.) That knowledge is important, since without knowing your Social Security benefit it’s hard to plan how much to save on your own or when to retire.
Sass reviews two research papers on the Statement, one that I wrote for RAND and a second by Giovanni Mastrobuoni, now at the University of Essex in the UK. Both papers used data from the Health and Retirement Study. The HRS first asks near-retirees what they think they’ll receive from Social Security, then later (usually two years) asks them what they actually received. Comparing predictions to actual benefits gives us an idea of how well Americans understand Social Security.
The two papers come to slightly differing conclusions: I didn’t find that wider distribution of the Social Security Statement over time has done much to improve near-retirees’ ability to predict their benefits. Mastrobuoni did find that, among people who didn’t visit a Social Security Administration office to get a personalized benefit estimate, receiving the Statement made them more likely to be able to produce an estimate of their benefit. These two results aren’t necessarily contradictory: it’s possible that after the introduction of the Statement fewer people contacted SSA to get a personalized estimate, so even improvements among these non-contacters might not raise overall knowledge. Likewise, Mastrobuoni focuses on whether near-retirees are willing to guess at their future benefit, while I focus on how accurate their guesses turn out to be.
But that misses what I now think is a much more important point: that even after years of receiving (and hopefully reading) the Social Security Statement, most people have very little idea what they’ll get from Social Security. As of 2008 (the final year of data in my paper) 30 percent of near-retirees couldn’t even make a guess as to what their benefit would be, even though most of them were less than two years away from retirement. Of the 70 percent who did make a guess, about half were off by more than 25 percent.
In other words, nearly two-thirds of near-retirees are unable to guess their Social Security benefit within plus/minus 25 percent. One-in-ten retirees received a Social Security benefit that was only half as much as they thought they would receive. That’s like having your 401(k) balance drop by 50 percent on the day you retire, with no hope for recovery.
To me, this is shockingly poor level of knowledge given that the point of a “defined benefit” pension like Social Security is to deliver benefits that – unlike a defined contribution 401(k) plan – can easily be known in advance. How are you supposed to decide how much to save in your 401(k) or when it’s safe to retire if you have little idea of the base income you’ll receive from Social Security?
The fault lies not in us or in the Statement, which does a reasonable job of projecting benefits, but in Social Security’s benefit formula, which is a lot more complicated than that of a traditional “final earnings” corporate pension plan. There are multiple benefit calculations and indexing procedures that make no sense to a typical person. If that person can’t understand his benefit even after having received the Social Security Statement, the problem is with how we’ve chosen to calculate Social Security benefits.
And much of this complexity is to no end. Social Security provides relatively weak poverty protections, since the truly poor – those with spotty work histories – often don’t qualify. Of the poorest fifth of Americans, about 20 percent won’t qualify for benefits in retirement, according to Social Security Administration projections. Social Security also provides poor work incentives for married women and people thinking of delaying retirement.
I’ve argued that Social Security reform – which is inevitable, given the program’s growing deficits – make simplicity a goal along with solvency and improved social protections. I favor shifting Social Security to a universal, flat benefit set at the poverty threshold, to insure that no one retires into poverty. That also happens to be a benefit that’s easy to understand, predict and plan around. On top of that, we need policies to expand individual retirement saving, such as making automatic enrollment into 401(k)s universal and expanding pension coverage by small employers by reducing the regulatory burdens of offering a 401(k). It may sound radical, but it’s not so different from the national retirement plans in the UK and New Zealand.
I’m skeptical of how well the Social Security Statement really works. But I don’t think the answer is a better-designed Statement so much as a better-designed Social Security program.
Also on Forbes:
