The mystery of the missing Social Security data

Health and Human Services Secretary Sylvia Burwell, center, flanked by Treasury Secretary and Managing Trustee Jacob J. Lew, left, and Labor Secretary Thomas E. Perez, release this year's Social Security Trustees Report. But something was missing.
(Susan Walsh / Associated Press)

Retirement expert Alicia Munnell of Boston College has brought a simmering Social Security mystery out into the open: Why were key figures about the program’s benefits for retirees deleted from the latest trustees’ report?

Munnell’s conclusion, as set forth in a column published this week on MarketWatch, is that the reason is nefarious.

“The deletion,” she writes, “is the culmination of a concerted effort by a band of critics” who contend that Social Security is too generous to retirees. Among their goals is to undermine a growing effort in Congress to increase and expand the program’s benefits, and make it easier for policymakers to cut back the benefits.

The deleted figures that concern Munnell measure the program’s “replacement rate.” This is the percentage of a retiree’s lifetime earnings replaced by Social Security benefits. It’s among the most important measures of the program’s effectiveness. The statistic has been listed in the annual Social Security Trustees Report since 1989 -- until this year.

For retirees with median lifetime earnings, the replacement rate has been somewhere in the neighborhood of 40% -- if, say, your lifetime average earnings were about $75,000, a Social Security benefit of $30,000 would give you a 40% replacement rate. (The calculations are somewhat more complex than that -- the replacement rate is designed to be higher for low-income workers and lower for wealthier workers -- but you get the gist.)


This matters because financial experts say that to retire without a jarring reduction in one’s standard of living requires a replacement rate of 70 to 75%. Social Security doesn’t aim to provide it all, but to supplement employer pensions and retirees’ savings in 401(k) plans and other accounts (both of which have gotten very shaky in recent years).

Until this year, a chart in the annual Social Security Trustees Report listed replacement rates for retirees at all levels of income, and projected them decades into the future. (See accompanying illustration.) This year, however, those figures disappeared from the report. In their place was a chart that showed benefits only in dollar terms. But in isolation, those figures don’t tell you anything about how those benefits correspond to retirees’ needs over time.

Deleting the replacement rates plays into the hands of those critics Munnell alludes to (but doesn’t name). Among them are the independent analyst Sylvester Schieber and Andrew Biggs of the American Enterprise Institute.

Their argument is that American retirees are, in fact, pretty flush. They try to show this in two ways. One is to suggest that estimates of retiree incomes systematically overlook a huge portion of the income that seniors live on. We analyzed this assertion in January, and found it profoundly misleading.

Their other point is that Social Security’s replacement rate formulas are faulty, and that the real rates are much higher than 40% -- more than 60% for average earners and 70% for low-income retirees, in fact. Part of their argument is that Social Security’s replacement formula gives retirees too much credit for wage growth in the economy, so the replacement rate seems lower than it really is.

As Munnell and Social Security’s own actuaries point out, some critics also base their replacement rate calculation on a retiree’s income in his or her last five years before taking Social Security. Since many retirees have no income in the last few years before starting benefits -- that’s 14% of all new retirees, the actuaries say -- the replacement rate for those years is infinity (that’s what you get when you divide a number by zero), which pushes up the overall replacement rate. The actuaries say the right way to calculate a replacement rate is against years with non-zero income.

In any case, if you can put over the idea that Social Security already is replacing close to 70% of retirees’ income, then you’ve damaged the argument that benefits should be made more generous.

But including zero-income years in the replacement rate calculation, Munnell says, “makes no sense.” Those last five years before retirement often are marred by poor health and disability, the twilight of a worker’s career -- they’re by no means reflective of his or her working experience.

The real question is why the replacement rates were deleted this year. The trustees didn’t explain the change, and until Munnell’s column, few people noticed it. But its impact on policy-making will eventually be felt.

“No rational case can be made for deleting replacement-rate numbers from the Trustees Report,” Munnell writes, and she’s right. The only reason is to clear the way for a concerted attack on the program and to undermine the campaign to improve it for the millions of Americans who depend upon it.

“If Social Security replaces less, then future workers must depend on what is now a fairly wobbly 401(k) system for more. Without replacement-rate numbers, policymakers will have no idea what they are doing to the retirement security of future workers as they consider alternative Social Security provisions.

“This is not inside baseball,” she concludes. “This is important.”

Keep up to date with The Economy Hub by following @hiltzikm.