42% of Americans Have No Retirement Number, Which Is Not the Same as Being Short

Forty-two percent of Americans have never calculated how much money they will need in retirement, according to a new report from Evergreen.ai. MarketWatch reported the figure on October 23, 2026. The story also ran on Morningstar's news service that morning, credited to Dow Jones and timestamped 6:00 a.m. The headline calls these savers headed toward retirement "blind." That word refers to the absence of a target, not to any measured shortfall in savings.

That distinction matters for anyone who reads the figure as a verdict on whether people are saving enough. Someone who has never run a retirement income projection can still have a 401(k), a pension, or a reasonable estimate of Social Security benefits. The 42% measures the missing target: the number a saver needs to turn a balance into a sustainable annual withdrawal, or the contribution rate needed from each paycheck. Without that target, savings decisions tend to be made by default. That is a different problem from being underfunded.

Other surveys point in the same direction, though they ask different questions. The 2026 Retirement Confidence Survey from the Employee Benefit Research Institute, fielded online from January 2 through January 28, 2026, found that 64% of Americans feel confident they have enough money to live comfortably throughout retirement. Confidence fell among both workers and retirees. The same survey found that 46% of workers have calculated how much they would likely need for retirement health expenses. Health costs are among the most concrete lines in a retirement budget, so nearly half of workers having a figure for that line, while 42% of Americans have no overall figure, is a useful contrast.

The Retirement Insecurity 2026 report adds a balance-sheet view. It reports that 80% of Americans said the nation faces a retirement crisis in 2026, up from 67% in 2020. It also finds that 47% of respondents have less than $100,000 saved for retirement, and 18% have no retirement savings at all.

Access is a structural constraint on all of this. Reuters, citing research from Economic Innovation Group, reported in February 2026 that nearly 54 million American workers have no access to any employer-provided retirement plan. Workplace plans are often where people first meet a projection tool or a contribution default, so a missing plan can mean the calculation never happens. This is an inference, but it is the mechanism most worth testing against the data.

The Employee Benefit Research Institute also runs its Retirement Security Projection Model, which simulates the share of the U.S. population at risk of falling short of the retirement income it will need. That is a different kind of evidence from asking people what they believe. A modeled estimate rests on assumed incomes, savings paths, and benefit formulas, and its output moves a great deal when those assumptions change. The practical question for any claim is which measure it rests on: self-reported confidence, self-reported calculation, a balance distribution, or a simulation.

In the 23rd Annual survey cited by the Transamerica Center, 41% of workers said future generations of retirees will be worse off than people retiring now. Goldman Sachs has also published its Retirement Survey and Insights Report 2026, based on a survey of working and retired Americans about the financial strain reshaping retirement.

For savers, the takeaway from the Evergreen.ai figure is narrower than the headline. Forty-two percent of Americans lack a stated retirement target, and that gap can be measured separately from the balance in any account. Whether it leads to lower outcomes depends on who the 42% are, what they have saved, and what benefits they expect. Anyone building a plan from these numbers should start with the target, because the calculation still has to be done household by household.