If you were a government agency that produced a study whose headline result was repeatedly misinterpreted, what would you do when you updated that study? Maybe hold back on the headlines that were misinterpreted? Not if you were the federal Government Accountability Office, which in an update to previous work declared this week that 48% of U.S. households aged 55 and over in 2016 “had no retirement savings.” That claim is factually incorrect, and the way it will be interpreted by politicians and the press will only make it less true.
Senator Bernie Sanders has repeatedly cited the GAO’s past results, rephrasing them in ways that are entirely rational for a non-specialist but which lead to radically incorrect results. A recently as January 2019 Sanders tweeted:
“Today, in America, more than half of older workers have no retirement savings – zero. That is unacceptable. We must create an economy that works for all of us, not just the wealthy few.”
But, as FactCheck.org and others have noted, this statistic – derived from the Federal Reserve’s Survey of Consumer Finances – excludes anyone who has only a traditional defined benefit pension. If we count both retirement accounts and traditional pensions, 72% of households aged 55 and over have retirement savings. That’s up from only 64% in 1989 – so things are getting better, not worse. But you wouldn’t know that from the press coverage.
Excluding traditional pensions from “retirement savings” defies common sense, and it isn’t even justifiable on a technical level. In the National Income and Product Accounts and the Federal Reserve’s Financial Accounts of the United States, household wealth includes both defined contribution retirement plan assets such as IRAs and 401(k)s and the future benefits owed to employees via traditional defined benefit pensions. I can’t see why the GAO’s headline number would categorize this group – most of whom are public sector employees with generous pensions – as “non-savers,” especially when that limited definition of retirement savers has already caused confusion.
And there’s a second issue that will cause readers to be misled: the distinction between households and individuals. The Fed’s SCF measures things on a household basis, whereas in public policy we mostly care about individuals. The issue comes in that married households or non-married couples – with two people – are much more likely to have retirement savings than unmarried, single-individual households. Among married households aged 55 and up, 83% had a retirement account or pension plan, versus only 59% of single-individual households. This means that about 77% of Americans 55 and up have retirement savings, versus only 72% of households. How do I know people will be misled? Because they already have been: for instance, Bloomberg’s Ben Steverman, an experienced financial reporter, repeatedly interpreted the GAO’s household figures to mean individuals.
So the true state of retirement savings – 77% of older Americans with retirement plans – is radically different than the GAO’s stated 52%. It’s a 25 percentage point difference, or a 48% relative difference in the number of older Americans with retirement savings. They’re simply very different numbers. And yet, I can see it now: almost no one who will cite the GAO’s 52% figure will be moved in the slightest by the correct figure of 77% of older Americans having a retirement plan. Just as they aren’t moved when they find that out the true median retirement income is 30% higher than is stated in the Current Population Survey, the dataset used to measure elderly poverty and reliance on Social Security. Or how they don’t care when it’s pointed out that 81% of full-time employees have access to a retirement plan at work, as shown in the Bureau of Labor Statistics’ National Compensation Survey, versus the sub-40% figure that retirement crisis proponents cite based on the faulty Current Population Survey data. Many people are utterly convinced that America faces a “retirement crisis” and no data, argument or evidence will convince them otherwise.
But let’s say you did care about the data. You might nevertheless point out that 23% of older Americans without a private retirement plan is a big problem, which is a perfectly legitimate claim. And yet being without a private retirement plan doesn’t mean you’re without retirement income. Even “non-savers” almost certainly have Social Security benefits, and some – perhaps small businessmen or farmers – have retirement income coming from outside a formal retirement plan.
To illustrate, let’s look at near-retirees in 2001 who didn’t have either a retirement account or a traditional pension. These households, aged 50 to 59, had median household earnings of just $23,688 in 2016 dollars. So this isn’t a rich group of people who you’d expect should be saving much for retirement outside of Social Security. But that gives you an idea of the level of pre-retirement earnings they’ll be looking to replace once they retire. Now fast forward 15 years to 2016, when this same age-group of people would now be aged 65 to 74. Their median household income in 2016 was $20,901, equal to 88% of their pre-retirement earnings – a more than adequate “replacement rate” by most financial advisors’ recommendations. In other words, even these non-savers are on average doing okay in retirement.
One might argue, a government report can’t contain this kind of detail and nuance. But why not? I ran these figures while sitting by the pool on vacation, so why can’t a government agency give elected official and the public the facts and data they need to make informed decisions on retirement policy?
