Andrew Biggs

Andrew Biggs, co-author of a recent piece questioning whether U.S. retirees are as poor as we thought, during a congressional hearing last month. (Mark Wilson/Getty Images)

That's the implication of an op-ed in Thursday's Wall Street Journal, which says that figures on retiree incomes used by the Social Security Administration systematically miss as much as 60% of the income that seniors receive. That income, they say, is IRA and pension (especially 401(k)) income.

The goal of the authors, independent benefits consultant Sylvester J. Schieber and American Enterprise Institute scholar Andrew G. Biggs, is to question the growing consensus by progressives that Social Security needs to be expanded to meet the needs of present and future retirees. The consensus is based on statistics showing that the program is the major source of income for most of the elderly -- at least 90% of the income for nearly half of unmarried retirees, for example -- and that the elderly are increasingly strapped as they enter retirement and grow much worse off as they age. The most prominent backer of this view is Sen. Elizabeth Warren (D-Mass).

If there's a supply of wealth and income that isn't being counted, of course, that's a different story. 

"We are certainly raising a question about those statistics," Schieber told me in an email. If the Social Security statistics aren't counting an important source of income, he said, "then it would seem that the share of income attributable to Social Security is likely exaggerated, especially relative to pensions/IRAs." Schieber lays out his case in greater detail in an article with a different co-author to be published shortly in the Journal of Retirement.

The question is whether the authors' findings hold true for most retirees. There's reason to believe they don't.

It's proper to observe that Schieber and Biggs are both longtime critics of Social Security. Biggs, who was a deputy commissioner of Social Security in the George W. Bush administration, has advocated privatizing the program and converting it to something resembling a means-tested welfare program. Schieber's record as a critic and skeptic of Social Security as it's currently organized dates back at least to the 1990s. But they're both serious students of the program, and their assertions should be judged on their own terms.

So let's take a look.

The core of the authors' argument is that Social Security relies on income data from the Census Bureau's annual Current Population Survey -- but the survey ignores most of the income Americans get from their 401(k) and IRA plans. That's because the survey counts only payments made on "a regular, periodic basis" -- not income from occasional, as-needed withdrawals from those other accounts.

The discrepancy is large in the aggregate, they observe: For 2008, the CPS reported that retirees collected $5.6 billion in individual IRA income and $222 billion from pensions and annuities, but retired households reported $111 billion in IRA income and $457 billion in pension and annuity income to the IRS.

One question about this finding is raised by progressive economist Dean Baker, who observes that other commonly used surveys of household income that do count the "missed" income sources come to roughly the same conclusions about the financial status of the elderly as the CPS.

Even if the elderly are collecting more income than the Social Security Administration counts, however, that still leaves a couple of important issues: How much richer are they than we thought, and who among them is really getting all this money?

That's where the Biggs and Schieber analysis may be most seriously flawed. Schieber explained to me that he defined retirees as people receiving Social Security benefits -- a rough approach, since it would include disabled but unretired people, but close enough. In 2008, that was 23.4 million "tax-filing units," or households.

But that means that, on average, we're talking about $4,513 per household in uncounted IRA income and $10,056 in pension and annuity income. That might be considered a significant bump up in income, measured against Social Security benefits that topped out at $26,220 a year in 2008. (The maximum is $31,704 this year.)

But of course IRA and pension income isn't evenly distributed across the income spectrum. Like most asset-based income, it clusters toward the top end of the lifetime income scale. Schieber acknowledges in his Journal of Retirement article that as much as 30% of Social Security recipients don't file with the IRS because they don't collect enough in outside income to make their Social Security benefits taxable.

These are almost certainly the poorest retirees; their pension and IRA income is sure to be minimal, not likely to amount to a significant share of total income. (Social Security benefits become taxable only if half the benefit plus outside income reaches $25,000 or more for singles, $32,000 for married couples.) Schieber's extended article shows that for those in lowest 10% to 20% of retirees, for whom the median Social Security benefit was $13,100 in 2008, the median uncounted IRA/pension income was about $1,000 a year. For the top 10%, however, the median uncounted amount was $44,760.

So the question remains, are American seniors really rich? The answer from Schieber's work is: For the most part, no. For a median retiree, his article states that the Social Security Administration counted his or her household income as $29,757 and the IRS data placed it at $34,530.

It's true, as he wrote there, that "official government reports on the income status of retirees ignore hundreds of billions of dollars of their income." But that doesn't support Biggs' and Schieber's implication that this means American retirees aren't facing real financial trouble, or that Social Security isn't still the hook on which the retirement security of most individual Americans depends.

Reach me at @hiltzikm on Twitter, Facebook, Google+ or by email.