Retirement in America: Were We Better Off 50 Years Ago?

Measuring Retirement Incomes

Over the years, I have encountered many articles wondering whether the prospects of American retirees have improved, but never one that directly compared the income received by retirees who lived in the past with the income received today. This column does just that.

Specifically, I calculate the median income for 1) 1973′s private-sector retirees and 2) 2021′s private-sector retirees. Evaluating the change in circumstances between two endpoints, at a single spot on the earnings distribution, greatly simplifies what would otherwise be a very long inquiry. (Also, as most retirement debates invoke the fate of typical Americans, the median income is a logical choice.)

For the same reason, I set aside public-sector employees. Their situation also warrants discussion, but not here.

Then (1973)

In 1973, the median monthly Social Security retirement check delivered the less-than-princely sum of $166 (tab C3 on the spreadsheet linked to “download all tables,”) That converts to $1,004 per month in 2021 dollars. Those retirees, of course, held no money in defined-contribution plans, as the 401(k) system had not yet been invented. A few profit-sharing plans existed, and some investors owned taxable accounts, but those assets were beyond the means of median-income employees and can safely be ignored.

Ordinary Americans, however, might have been served by a pension plan. If so, their benefits would likely have been substantial. On average (for this figure, I do not have access to the median), retirees who qualified for private-company pensions in 1973 received $177 per month—almost exactly the median payment supplied by Social Security. Effectively, their pension receipts doubled their pay.

That’s an excellent haul! Unfortunately, the pension system was not comprehensive. Most small and many midsize companies did not offer pension plans. Also, they were primarily designed for long-term employees. Workers who often changed jobs, as with my father (this topic hits very close to home), rarely qualified. In 1973, 44% of private-sector retirees received pension income, meaning that the majority did not.

Calculating expected income for the median-income retiree therefore requires multiplying that $177 check by 0.44, for a percentage-weighted median of $78 ($471 in 2021 dollars). The percentage-weighted approach is imperfect, as it understates the wealth of employees who held pensions while simultaneously overstating the income of the remaining 56%. That’s still better than pretending that private-sector pensions were either ubiquitous or nonexistent.

The table below summarizes the results. Expressed in 2021 dollars, the median private-sector retiree of 1973 received $1,004 in monthly income from Social Security, plus $471 from a pension, for a total of $1,475.

Median Private-Sector Retiree: 1973 (Monthly Income $)

Now (2021)

Social Security benefits have increased over the past half-century. They track real incomes, and real incomes across all wealth percentiles (Table H-3) have risen. Thus, the median monthly Social Security payment in 2021 was $1,658 (the same source as before), a 65% increase from 1973′s level, after adjusting for inflation. Through that boost alone, current retirees are wealthier than their predecessors.

Just as well, because their pension income has withered. As is well known, most private-sector pensions have vanished, leaving only 11% of corporate employees participating in such plans. Less appreciated has been the decline in benefits for those who do qualify. At $177, the average monthly pension payout of 1973 represented $1,071 in 2021 dollars. By 2021, that amount had shrunk to $884.

With pensions all but disappearing, defined-contribution plans have attempted to fill the gap. At that task, they have not fully succeeded. Although I was unable to obtain the industrywide median 401(k) balance for retirees, I did find the figure for Vanguard’s plans in 2021: $87,725 (Page 47). That is clearly insufficient. Assuming a 5% withdrawal rate, that translates to monthly income of $366, far below the median benefit paid by 1973′s pension plans.

It does no good to protest that only about half of 1973′s retirees received pension income, because the same caveat applies to defined-contribution assets. Currently, 69% of full-time private-sector employees have access to a defined-contribution plan, and about three fourths who do have joined that plan. The overall participation rate is thus a modest 52%. On a percentage-weighted basis, in 2021 the median monthly payout from defined-contribution assets for retirees was a distinctly unimpressive $190.

Summing the monthly proceeds from the three sources of Social Security, pensions, and defined-contribution plans makes for a total of $1,945, for the median-income 2021 retiree.

Median Private-Sector Retiree: 2021 (Monthly Income $)

Conclusion

Given my family background—among my grandparents, parents, uncles, and aunts, only one ever received a corporate pension—I have long distrusted nostalgia about the golden age of retirement, when everyday workers could allegedly retire without financial worry. There certainly were employees who fit that description, but that narrative ignores the tens of millions who did not.

It must be confessed, however, that while today’s median private-sector retirees are undeniably better off than their forebears, their progress owes not to the development of defined-contribution plans, but instead to the quiet increase in Social Security payments. Retirees’ boats have floated on the tide of overall economic growth rather than on improvement to the retirement system itself.

For defined-contribution plans to improve upon what the pension system once provided, rather than a second-rate substitute, two things must occur. One, they must become universal. Every company must offer a plan—and all employees must participate unless they opt out. (Forced compliance would be a step too far.) Second, contribution rates need to improve. They need not be exorbitant, as a 7% lifetime rate should suffice to create a pensionlike payout. But they must exceed their historic levels.

Finally, if there remains any doubt that today’s median-income retirees have the advantage on 1973′s vintage, consider this: Few of those pension-plan checks contained full cost-of-living increases, and many had none at all. Indeed, concluded a government study, from 1973 through 1979 the average pension benefit grew 40% as quickly as inflation. That is, while prices were up 63%, pension payments gained only 25%.

Sometimes, the good old days were not.