Experts disagree on how well prepared Americans are for retirement. Data on income and savings aren’t perfect, and economists differ as to how much they think a household actually needs to fund a comfortable retirement. Some studies find a large retirement-savings gap, while others — including some well-respected work — find that most Americans are doing well in saving for retirement. I’m on the more optimistic side of that debate, but any honest researcher would admit that a legitimate debate on retirement income security exists.
Which is why the Washington Post’s lengthy exposition of “The New Reality of Old Age in America” is so disheartening. It has everything that poor journalism exhibits: excessive reliance on anecdotes, the use of misleading and sometimes outright incorrect data, and the one-sided citation of experts.
While some work by choice rather than need, millions of others are entering their golden years with alarmingly fragile finances. Fundamental changes in the U.S. retirement system have shifted responsibility for saving from the employer to the worker, exacerbating the nation’s rich-poor divide. Two recent recessions devastated personal savings. And at a time when 10,000 baby boomers are turning 65 every day, Social Security benefits have lost about a third of their purchasing power since 2000.
Nearly all of this is either misleading or simply wrong.
Americans are in fact working longer, which is both to be expected as lifespans increase and to be celebrated as seniors overcome age discrimination, health factors, and the deterioration of their technical skills to extend their work lives and build retirement savings. Moreover, as I pointed out in a 2015 Wall Street Journal op-ed, retirees aren’t all working at Walmart-greeter wages. Older Americans’ labor-force participation rose even as that of younger Americans fell, but older workers maintained their wage levels.
So the Post’s whole narrative that working longer is evidence of a retirement crisis is suspect. Worse, most of the Post’s supporting evidence also falls short.
Just to start, one might expect that if retirees had “alarmingly fragile finances,” as the Post claims, they would also have a high poverty rate. And yet we see precisely the opposite: Recent research from two Census Bureau economists, C. Adam Bee and Joshua Mitchell, finds that only 6.9 percent of Americans aged 65 and older have incomes below the poverty line. The poverty rate among working-age Americans is roughly double that level; among children, it’s at least triple. If poverty drops as Americans retire, how can that be a sign that our retirement system — which combines Social Security with employer-sponsored retirement plans and personal saving — isn’t working?
The Post cites left-leaning economist Teresa Ghilarducci, who states:
There is no part of the country where the majority of middle-class older workers have adequate retirement savings to maintain their standard of living in their retirement. People are coming into retirement with a lot more anxiety and a lot less buying power.
Really? The Census Bureau’s Bee and Mitchell find that, from 1989 to 2007, the inflation-adjusted income of retirees at the 25th percentile of the income distribution — precisely the types of working-class households the Post highlights — rose by 50 percent. Is that a “lot less buying power”? Certainly not compared with the incomes of pre-retirees: Among the households of Americans aged 50 to 60, incomes rose by only 15 percent from 1989 to 2007, according to Survey of Consumer Finances data that I pulled up.
And that increase in retirement incomes is being driven by the private retirement plans that the Post and the critics it cites claim are failing. For the typical new-retiree household, the real value of Social Security benefits increased by 25 percent from 1989 to 2007. But the value of benefits from private retirement plans — meaning traditional pensions, IRAs, and increasingly 401(k)s and similar plans — rose by 75 percent. In light of such statistics, it is preposterous to claim that America’s private retirement-savings system is failing.
The Post also errs in claiming that “Social Security has lost a third of its buying power” since 2000, a claim the paper sources to a “study by the Senior Citizens League.” The Senior Citizens League study does not calculate changes in prices but changes in household expenditures, which encompass both rising prices and increases in the quantity of goods and services purchased by retirees. Why not instead cite the Bureau of Labor Statistics, which states that an inflation adjustment called the chain-weighted CPI — which would show that the buying power of Social Security has increased over time — is “a closer approximation to a cost-of-living index than other CPI measures.” (The Senior Citizens League study is based on a version of the CPI called the CPI-E.)
Likewise, a Post graphic trumpets that “Social Security payments are less than the federal minimum wage,” but produces this result only by using misleading statistics. The average annual Social Security retirement benefit is $1,369 per month, or $16,428 annually. A person working at the $7.25 federal minimum wage for 40 hours per week, 52 weeks of the year, would receive $15,080. The Post claims lower average Social Security benefits by looking only at benefits for “retired workers with an aged spouse,” which means a spouse who had less than ten years of work and thus doesn’t even qualify for benefits on her own. This group of retired workers with aged spouses is uncommon, making up only around 11 percent of Social Security retiree beneficiaries. Moreover, it is misleading to compare such a couple’s Social Security benefits to what they could have earned if both worked at the minimum wage, since an aged spouse beneficiary has less than ten years of work at any wage. The Post graphic also misstates the poverty threshold for an over-65 couple, which is not $16,242 but closer to $14,604. It is worrying that the Post article’s authors were willing to undertake such statistical gymnastics to demonstrate a point that is essentially false.
The Post article also claims that 33 percent of retirees receive at least 90 percent of their income from Social Security. But these figures are mistaken; they are derived from a household survey that, as shown in the Census study cited above, dramatically undercounts benefits from private retirement plans. IRS tax data show that only 18 percent of retirees receive 90 percent of their income from Social Security benefits.
The Post authors likewise call 401(k)s a “Wall Street gold mine,” despite the fact that the average fee on a 401(k) account has dropped by 38 percent since 2000 while investment fees paid by traditional defined-benefit pensions — particularly in government — have skyrocketed as these plans chase returns through exotic, high-cost “alternative investments” such as hedge funds.
Similarly, the Post authors cite data showing that “nearly 30 percent of households headed by someone 55 or older have neither a pension nor any retirement savings.” But households without retirement savings are also among the poorest, with average household earnings among near-retirees of only around $13,500 per person. It’s unrealistic to expect they’d have much saved, which is precisely why Social Security provides a progressive benefit for low-income households.
The American retirement system is far from perfect, but a lot of things are going in the right direction.
In discussing the shift from traditional pensions to defined-contribution 401(k)-style plans — a shift that is taking place around the globe — the Post authors claim that 401(k)s “work best for the wealthy.” Maybe so, but that doesn’t mean 401(k)s are worse than traditional pensions. Social Security Administration data for 1980, near the peak of traditional pension coverage, found that fewer than one-in-ten new retirees received any benefits from a private pension; even among the richest quarter of the population, only half received private pension benefits. Today, research from Investment Company Institute and IRS researchers shows that 81 percent of retiree households receive at least some benefits from private retirement plans.
The authors of the Post article might respond that I’m citing some relatively obscure data and studies here, where they relied on prominent statistics and reports. But that’s why a good reporter will access expert sources from a variety of perspectives, to find out which statistics make sense and which might be misleading or outright incorrect. As it stands, the Post authors cite only sources that agree with their gloom-and-doom view of the U.S. retirement system, left-leaning sources like Teresa Ghilarducci of the New School and the union-affiliated Economic Policy Institute.
None of this is to say that the retirees highlighted by the Post don’t deserve both our sympathy and the concern of policymakers. But in a news articles, individual examples should be chosen to illustrate broader points that are supported by data. And the Post doesn’t do that. The American retirement system is far from perfect, but a lot of things are going in the right direction — including for the low-income Americans that the Post focuses on. The article’s authors seem ignorant of nearly all of them.
— Andrew G. Biggs is a resident scholar at the American Enterprise Institute. He formerly served as the principal deputy commissioner and the deputy commissioner for policy at the Social Security Administration.