In fact, the retirement crisis narrative has become so pervasive that stakeholders willing to challenge the conventional wisdom are few and far between.

Enter Mr. Biggs, who's been pushing back against the narrative for several years. His efforts have redoubled since the release of the GAO's report, which Biggs charges is based on incomplete, cherry-picked data that fails to stand up to scrutiny.

"It comes down to differences in how you weigh the evidence," said Biggs, who commended the GAO's previous work on Social Security. "The report is based on this built-in narrative that the country is facing a retirement crisis. They're grasping on to evidence that supports that point of view, and not paying enough attention to evidence that opposes it."

He acknowledges that many of the facts in the 150-plus-page report are true, but also says some are provably false.

In the report's preface, the GAO cites data from the Social Security Administration showing 34 percent of retirees relied on Social Security for 90 percent of their income in 2015.

The SSA bases that figure on the Census Bureau's Current Population Survey. But a working paper from two economists at the Census Bureau, published last summer, shows that survey, which measures household income and poverty rates, dramatically underreports income from defined benefit pensions and retirement accounts.

Using more complete income data from the IRS, the paper found that the actual number of retirees relying on Social Security for 90 percent of their income is 18 percent—nearly half the figure cited by the GAO.

"They should know the old number is wrong," said Biggs. "Did nobody point that out?" More than 25 GAO staff members contributed to the report.

Charles Jeszeck, director of GAO's Education Workforce and Income Security Team, which produced the report, said the Social Security data came from published agency reports, and noted that the new numbers from the Census Bureau are not official, and are part of a report that isn't yet complete.

"Every data source has imperfections or degrees of uncertainty, and we welcome emerging work in this area," Jeszeck wrote in an email. "Regardless, whether you use the 34 percent cited in our report or the 18 percent as estimated [in the new Census working paper], that still leaves millions of people over age 65 relying on Social Security for 90 percent or more of their income."

Jeszeck said the GAO's recommendation for a Congressional commission, and the overall report, are not reliant on the specific Social Security data point in question.

"We stand by the findings and conclusions of our recent report on the nation's retirement system," added Jeszeck.

Not counting defined benefits

Some of the facts in the GAO's reports on retirement are startling.

A 2015 study by the GAO, compiled at the request of Sen. Bernie Sanders, I-VT, said 41 percent of pre-retirees had no retirement savings. The median household income of that cohort was $26,000, according to data from the Federal Reserve.

But to understand the full impact of that statistic, it's necessary to account for the fact that the GAO does not count income from defined benefit pensions as retirement savings—only defined contribution and IRA accounts are counted.

In fact, the GAO's report shows that 32 percent of the pre-retirees who reportedly have no retirement savings actually have a defined benefit plan.

"We broke them (DB plans) out as a separate component because they are rapidly disappearing in the private sector," explained Mr. Jeszeck. "Benefits from DB plans are certainly an important component of retirement income. In the future, fewer and fewer workers will have such benefits and so retirement saving in the form of DC plans and IRAs will assume an increasingly prominent role."

The actual number of pre-retirees with no form of retirement savings—401(k), IRA or defined benefit pension–is 27 percent, GAO notes in the 2015 report. Asked if discounting pension benefits from retirement savings could skew the question of whether a retirement crisis is impending, Jeszeck noted that 27 percent "is still a number in the millions."

Relying on incomplete data

When GAO refers to Census Bureau figures, Biggs claims the agency is relying on data that doesn't count all income from retirement plans.

"Once you see their factoids, and accept them as true, it shapes how people think about retirement security," said Biggs

GAO also uses the Survey of Consumer Finances, a triennial survey sponsored by the Federal Reserve Board, for much of its data.

That survey does a better job of capturing retirement assets than the Census Bureau's Current Population Survey, but even the Federal Reserve's survey fails to account for about 15 percent of what IRS data shows to be median retiree income, says Biggs. And that further calls into question some of GAO's core findings.

Biggs says the data on today's retirees shows they are not suffering a retirement crisis.

"The vast majority of current retirees are doing just fine, and they will tell you that," he said. "But if you looked at them when they were on the verge of retirement, they looked just as unprepared as today's pre-retirees do."

To prove that, Biggs recently compared near-retirees without retirement savings in 2007 to the reported income of households age 64 to 70 in 2016.

In 2007, non-savers had average household earnings of $22,290, according to the Survey of Consumer Finances, and total retirement incomes of $42,962.

By 2016, when those pre-retirees were at or beyond retirement age, their average household income was $39,703. Moreover, Biggs cites SCF data showing the non-saver retiree received only $13,951 in Social Security benefits in 2016, meaning they had substantial sources of retirement income outside of Social Security.

That data shows that those counted as "non-savers" in 2007 are much better off as they enter retirement today, and have managed to replace more than 90 percent of their working income in retirement, according to Biggs' analysis.

Nostalgia for DB pensions

The newer data from the Census Bureau—produced in the paper that Mr. Jeszeck notes is not completed—revised median household income for retirees to $44,400, up from the $33,800 previously calculated. The poverty rate among households older than 65 was revised to 6.9 percent, down from the 9.1 percent previously calculated.

"The discrepancy is mainly attributable to underreporting of retirement income from defined benefit pensions and retirement account withdrawals," according to the Census Bureau's working paper.

To be clear—that study is the independent work of two economists at the Census Bureau, and not considered official agency research.

There is also the question of the relevance of today's retiree income to what tomorrow's retirees will experience.

"The authors themselves have made it clear in both their working paper and in presentations that their findings cannot easily be extrapolated to future generations," said Jeszeck.

Biggs acknowledges that measuring the experience of today's retirees to tomorrow's is not necessarily an apples-to-apples comparison.

"It's not perfect, but it helps to look at how today's retirees are faring," said Biggs.

Biggs also claims GAO's most recent report is overly dependent on what he calls the "nostalgia" for private sector defined benefit plans.

According to data from the Labor Department, there were about 25.5 million participants in roughly 18,000 single-employer defined benefit plans in 1975—not counting multiemployer plans or government-sponsored plans.

In 2014, the number was about 27.6 million in about 7,650 plans.

Participation in DB plans peaked at 38 percent of the private-sector workforce in 1975, according to research from the former director of the PBGC.

While the number of plans has greatly diminished over the decades, the notion that corporate pensions were once enjoyed by most is false, argues Biggs. What's more, he says the history of participation numbers fails to tell the extent of pension uniqueness in the past. Due in part to strict vesting requirements, only one in 10 traditional pension participants actually collected benefits. Research on new retirees in 1980 shows pensions were centralized in the wealthiest quarter, with about half of them receiving pension income.

"GAO's nostalgia for corporate defined benefit plans is not backed by the data," said Biggs.

Biggs has also taken aim at how the GAO uses diminished savings rates to impute inadequate retirement savings, and questions the thoroughness of reports the agency uses to show America's alleged lackluster standing on retirement security compared to other developed countries.

"I'm not trying to be too hard on them," Biggs said of his criticisms of GAO's retirement work. "There is a lot of work in their reports that is correct."

Nor is Biggs against retirement reforms. He favors auto-enrollment requirements for existing sponsors of 401(k) plans, and he advised Sen. Marco Rubio, R-FL, on a plan that would open participation in the Federal Thrift Savings Plan to private sector workers without access to a workplace savings plan.

And he is not against new backstop programs to address those who clearly do fall through the cracks. A former deputy commissioner of the Social Security Administration, Biggs has advanced reforms that would increase benefits for the neediest Americans, and reduce benefits for the wealthiest.

But he definitely believes GAO's work is skewed to the assumption that a retirement crisis is coming, and he worries that the trade and business media have blindly accepted the prospect as gospel.

"There is too much advocacy for the conventional wisdom," said Biggs. "One really useful role an agency like the GAO could play would be to stress test that conventional wisdom."

Read the GAO's response at The Hill.com to Biggs' criticisms.

Read Biggs's criticisms of the GAO reports here and here.

Read the 2015 GAO report.

Read the 2017 GAO report.