Americans worry about retirement, and understandably so.
Not because there’s an actual retirement crisis in the making, but because Americans are everywhere told that there is.
A recent lengthy Wall Street Journal treatment of Americans’ retirement savings – titled “A Generation of Americans Is Entering Old Age the Least Prepared in Decades” – is one of the most depressing examples I have seen recently, and I’ve seen many. The Journal’s coverage of Americans’ retirement prospects serves up flawed data tied to touching but unrepresentative personal case studies, all the while ignoring clear indications that retirement savings and retirement incomes have never been higher.
Saving for retirement, in simplest terms, is about being able to maintain your pre-retirement standard of living once you cease working. So how are we doing? It depends on who you ask. There is an entire cottage industry, an odd alliance between progressives who wish to expand Social Security and financial industry executives who want you to buy more of their saving products, who argue that most Americans are saving well short of what they need to provide a decent lifestyle in retirement.
The Journal’s exposition wholeheartedly embraces this viewpoint, while lacking even the basic journalistic responsibility to double-check their facts. As a result, the Journal’s narrative is brutally undercut when accurate data are presented.
A central claim of the Journal article is that retirees’ “median incomes including Social Security and retirement-fund receipts haven’t risen in years…”
If retirees’ incomes truly haven’t risen since 2000, as the Journal claims, this would be a problem: for retirees to maintain their pre-retirement standards of living, their retirement incomes should rise at roughly the same rate as the earnings of workers in the economy.
And yet, the Journal’s claim of stagnant retirement incomes is simply and unequivocally false. In reality, retirement incomes – for rich and poor alike – have risen substantially faster than the wages and salaries that retirement benefits must replace once individuals stop working.
The Journal produced their depressing retirement incomes figures by relying on a government dataset – the CPS – which has been proven to dramatically undercount the income retirees’ receive from private retirement plans, measuring less than half of total private pension payouts. A recent Census Bureau study which compared CPS figures to internal IRS tax return data found that the CPS understates the median retiree’s true income by a massive 30%.
The data cited by the Journal also understates the growth of retirement incomes: in contrast, the Census Bureau study found that the median retiree’s income rose by 32% above inflation from 1990 through 2012, a period during which SSA data show that real median wages for workers rose by only about 11%. When retirement incomes grow faster than worker earnings, retirees will generally have a greater ability to maintain their pre-retirement standard of living.
And incomes did not grow only at the median: over that same 1990-2012 period, incomes for lower-income retirees rose by 31% and the poverty rate among Americans 65 and older dropped from 9.7% to 6.7%, reflecting rising incomes even for poorest retirees.
Likewise, incomes increased more for younger retirees, aged 65 to 74, than for the oldest retirees aged 85 and over.
This is the opposite of what you would expect if the gradual shift from traditional defined benefit pensions to 401(k)s was, as the Journal reporters claim, undermining retirement income adequacy.
Much of this increase in retirement incomes is due to rising benefits paid from private retirement plans.
For instance, publicly-available IRS data show that from 1996 through 2016, total payouts from private retirement plans increased by 136% above inflation, versus an 80% real increase in total Social Security retirement benefit payments.
When other sources of retirement income grow faster than Social Security benefits, this implies that retirees are becoming less dependent on Social Security to get by in retirement.
Private pension benefits are also being more broadly shared. Again from 1996 to 2016, the number of Social Security retirement beneficiaries rose by 32%; over that same period, the number of individuals receiving taxable pension/annuity payments rose by 46% and the number of individuals receiving taxable payouts from retirement accounts such as IRAs and 401(k)s rose by 140%.
Hard data show that more Americans are collecting higher private retirement plan benefits and becoming less dependent on Social Security.
And yet, the Journal tells us, “Americans are reaching retirement age in worse financial shape than the prior generation.”
The Journal also hits on health costs in retirement, which are, to be sure, a matter of concern. But rising health costs are a matter of much bigger concern to the states and federal governments, which foot most of the bill for Medicare and Medicaid, than for retirees themselves.
Yes, examples of extraordinary health costs can easily be found; we’re a country of 300 million people with a highly-flawed healthcare system, so you’re going to see some very difficult individual cases. But again, there is no broad crisis.
Data from the Consumer Expenditure Survey show that out-of-pocket health costs are about the same percentage of retirees’ incomes today as during the 1980s.
While health costs have risen, retirees’ incomes have risen at least as fast. And, like the CPS, the Consumer Expenditure Survey tends to underestimate retiree’s incomes, so it is likely that true health costs have declined as a percentage of retirement incomes over the past several decades.
Other research on long-term care costs from economists at the RAND Corporation finds that – in good part to due to Medicaid and Medicare footing the bill – the average retiree household spends only $7,300 on long-term care over its entire retirement.
95% of households spend less than $47,000 on total long-term care costs throughout retirement.
Long-term care insurance makes sense for many retirees, but the fact that bankruptcies are lower among retirees than among working-age households tells us that truly catastrophic health costs are relatively rate.
The Journal is right to highlight the increase in households entering retirement with debt, which in part is a function of government efforts to encourage home and education lending.
But again, context is needed. The Federal Reserve’s Survey of Consumer Finances shows that, even including debt payments, most retirees do not even consume their entire incomes.
Likewise, for many households net worth inclusive of debt increases as they grow older.
One reason for the financial security of today’s retirees is that retirement savings have been rising rapidly.
Fed data show that in 1975, the height of traditional pension participation, total retirement plan assets were equal to just 50% of the wages and salaries those assets would replace in retirement. By 2015, total retirement savings topped 300% of total wages, a six-fold increase. Likewise, DOL data show that contributions to private retirement plans rose from an average of about 6% of employee wages in the 1970s to over 8% today, a shift that results in a roughly 25% increase in savings available to fund retirement incomes.
All of these facts help explain why 78% of current retirees tell Gallup they have sufficient money to live comfortably, versus only about 63% of working age households. Polls show that Americans worry about retirement, but once they reach it many of those worries go away.
There simply is not an income crisis among today’s retirees. The great majority of current retirees are doing fine.
Oh, but it’s not about today’s retirees, I sometimes hear.
It’s about futureretirees. I can hear Little Orphan Annie singing, tomorrow is always a day a way. Yet it's not clear why the crisis should be waiting for tomorrow. Participation in traditional defined benefit pensions – the decline of which the Journal calls “the biggest reason… many Americans [are] facing a less secure retirement than their parents” – peaked in 1975 at less than 40% of the workforce.
By the Journal’s logic we surely should have a retirement crisis already. Instead, retirement incomes are rising and poverty in old age is dropping.
If we’re interested in tomorrow’s retirees, why not consult with groups like the Social Security Administration and the Urban Institute, who have constructed the most sophisticated computer forecasts of retirement saving? Maybe it’s because these models don’t project a retirement crisis either.
For instance, the Urban Institute’s Dynasim model estimates that the median retiree in 2015 had income sources and assets sufficient to support a total annual income of $37,887. By 2025 real median incomes are projected to rise to $40,880, and to $42,165 by 2035. The Urban Institute model also projects that poverty rates in old age will fall, a reflection of high real incomes among the poor.
The Journal gets around this by instead citing the Center for Retirement Research’s (CRR) “National Retirement Risk Index,” which claims to show that 44% of Americans aged 55 to 59 have insufficient savings to maintain their pre-retirement standard of living. I like and respect the researchers at the CRR, but I just don’t buy these figures. For one, their model hasn’t been peer-reviewed and isn’t documented in enough detail for other researchers to replicate or disprove their results.
Reporters are happy to simply be given a number, so long as that number supports their story's conclusions, but researchers want to know how that number was generated. I’ve elsewhere raised a number of methodological concerns about the Retirement Risk model, most of which would bias it toward projection too many households at risk of undersaving.
Regardless, there are other studies, which are replicable and which have been peer-reviewed, which find that most households are prepared for retirement and retirement saving shortfalls, where they exist, tend to be modest. Why not cite them all so readers at least know there's a controversy?
The individual profiles the Journal includes aren’t much better. I understand that even a newspaper like the Wall Street Journal needs to highlight real-world stories rather than merely dry statistics.
But, in a responsibly-written article, individual examples are chosen to illustrate trends found in the data. At least one example included by the Journal is precisely the opposite.
The Journal highlights Arthur Smith, age 61, who after 31 years of contributing to a 401(k) ended up with only $45,000 in savings, half of which he withdrew to pay for a house in his late 50s.
Smith attributes his low balance to his 401(k) plan, which allowed him to invest in individual stocks rather than limiting him to diversified mutual funds. Investing in several high-risk stocks cost Smith half his savings.
All this is presumably true.
And yet, only a miniscule percentage of employees participate in self-directed 401(k)s that allow them to purchase individual stocks.
According to Vanguard, less than one-third of 401(k) plans allow for individual stock purchases and only 1% of employees offered a self-directed 401(k) use that option.
So the Journal’s supposed illustration of today’s retirement crisis in fact represents an uncommonly bad outcome among less than one half of one percent of 401(k) participants.
And of that tiny group, only a tiny proportion invest most of their 401(k) using that brokerage option; most keep the majority of their accounts in ordinary investments.
In reality, it is far more common for employees to take precisely the opposite approach: Vanguard reports that 58% of its 401(k) participants hold their savings entirely in target-date funds, which are broadly diversified low-cost mutual funds that automatically shift from stocks to bonds as the worker approaches retirement.
Others supplement target-date funds with additional mutual funds. Instead of day-trading, most retirement saving is becoming highly automated.
The Journal’s case-study, however tragic, is a black swan, not a trend.
Should the Journal’s reporters have known about all this? Sure, particularly had they read their own newspaper, in which I’ve publicized many of these findings over the years, including in a 2017 Wall Street Journal debate on retirement saving curated by one of the authors of the recent Journal piece.
Even if they hadn't seen any of this work, good journalism involves seeking out opposing opinions.
The Journal’s isn’t the first long-form piece on retirement savings adequacy I’ve had to take to task, and it surely won’t be the last.
Retirement is an extremely important issue, both to individuals making financial planning decisions and to policymakers deciding how to fix Social Security and set tax incentives for retirement saving.
A yet a great deal of media coverage on retirement saving is biased, financially illiterate or simply lazy. In short, much of the media’s coverage of retirement saving issues just stinks. That’s a great disservice to the public the media ostensibly exists to serve.
