Costco contributes 4% of pay to every worker’s 401(k) — no employee contribution required — rising to 9% for 25-year veterans. Boeing puts in 10%. Altria’s workers get 13% to 17% when profit-sharing lands. Stewart’s Shops reports over 200 cashiers who became millionaires through stock ownership. A national business publication calls this “jealousy-inducing.” The word you want is exhibit.
These numbers are real. They are also functionally meaningless to roughly half of private-sector workers who have no workplace retirement plan at all. The article profiles employers in a contest to outbid each other for talent at the top of the labor market. It does not mention the Bureau of Labor Statistics data showing that in the bottom wage quintile, employer-sponsored retirement-plan coverage hovers around 25%. The omission is structural, not incidental. A story about who gets 10% employer contributions that never mentions who gets zero is not personal-finance journalism. It is employer brand management, published at national scale with the institutional credibility attached.
The mechanism hiding in plain sight is the defined-contribution system itself. The 401(k), born as a tax-code accident in 1978 — Section 401(k) of the Internal Revenue Code, originally a deferral provision nobody intended as a primary retirement vehicle — replaced defined-benefit pensions for most new hires at large employers over the following three decades. Ford and GM’s 10% nonelective contribution the article celebrates is the residue of a bargain struck in 2023 when the United Auto Workers negotiated the number up from 6.4%. What the article does not say is that both automakers eliminated pensions for new hires roughly twenty years ago. The 10% nonelective contribution is not a gift. It is a downgrade the union fought to make slightly less catastrophic. A pension that paid a guaranteed benefit for life was replaced by a contribution into a market-dependent account whose terminal value depends on what the stock market does between the year you start contributing and the year you retire. The shift transferred investment risk from the employer — which had actuarial staff, a time horizon measured in decades, and the balance sheet to absorb volatility — to the individual worker, who has none of those things.
The defined-contribution system is, by design, a wealth-amplification machine for people who already have money. The tax deduction for 401(k) contributions is worth more to a household in the 35% marginal bracket than to one in the 12% bracket — the same $23,500 annual deferral (the 2025 limit for workers under 50) saves $8,225 in federal income tax for the top-bracket contributor and $2,820 for the bottom-bracket one. Workers who cannot afford to contribute — because they are servicing debt, covering childcare, or simply earning too little to divert cash from rent — forgo both the employer match and the tax benefit. The Employee Benefit Research Institute estimates that among workers earning below $30,000, fewer than 40% participate in an available 401(k). The Costco cashier who became a millionaire through disciplined contributions over decades is not proof the system works. The cashier is proof that the system requires a specific employer, a specific tenure, and a specific life history — no early withdrawals for medical emergencies, no layoff gaps, no divorce — to deliver the outcome the article holds up as normal.
The Secure 2.0 Act provision allowing employer matching for student-loan payments — the article names Boeing, Verizon, Chipotle, Comcast, Walgreens, and News Corp as adopters — deserves a closer look than the article gives it. The provision, enacted as Section 110 of the SECURE 2.0 Act of 2022 (Division T of the Consolidated Appropriations Act, 2023, P.L. 117-328), lets an employer treat qualified student-loan payments as elective deferrals for matching purposes. The framing in the article is pure employer-brand: look how innovative these companies are. The structural reading is that employers are now matching workers’ debt payments rather than wages — routing retirement benefits through a debt-service channel because the worker’s compensation is too low to generate both debt service and retirement savings simultaneously. A Boeing worker who puts 10% of salary toward a qualified loan gets the company match in their 401(k). That the worker needs to service the loan before saving for retirement is not addressed as a policy problem. It is packaged as a benefit.
The Vanguard data the article cites — average employer match of 4.7% of eligible salary, only 6% of plans offering 7% or above — tells the real story. The distribution is heavily right-skewed. The Costco and Boeing and Altria numbers are outliers at the far tail. The modal American worker’s retirement plan, if they have one at all, delivers less than half the employer contribution the article’s exemplars receive. And the 38% of the private-sector workforce that has no employer-sponsored retirement plan — not 401(k), not pension, not ESOP — is absent from the article entirely. Their retirement plan is Social Security, which pays a maximum benefit of about $58,400 per year in 2025 for someone retiring at 70, and an average benefit of roughly $23,000.
The publication of this article follows a familiar editorial architecture: profile the exception, let the reader generalize to the rule, and avoid naming the rule. The rule is that the American retirement system is a three-legged stool — Social Security, employer pensions, and personal savings — in which the second leg was sawed off and replaced with a tax-advantaged savings vehicle that delivers its largest benefits to workers who need them least. The ESOP model at Publix and Stewart’s Shops is the closest thing to the old pension bargain — employer-funded, long-horizon, not dependent on the worker’s individual contribution capacity — and it is available to a vanishingly small fraction of the workforce.
What the documentary record says is plain. The Pension Benefit Guaranty Corporation’s Pension Insurance Data Tables show that the number of private-sector defined-benefit plans fell from approximately 112,000 in 1985 — the documented peak — to a fraction of that number by 2020. The Federal Reserve’s Survey of Consumer Finances shows that median retirement-account balances for households headed by someone aged 55–64 were $185,000 in the 2022 survey — a figure that, converted to an annuity at conservative 2025 income-annuity pricing rates (roughly 5–6% payout), produces approximately $9,000 to $12,000 per year in income. The National Institute on Retirement Security estimated the retirement savings shortfall across all American households at approximately $7.1 trillion as of 2023.
Against that backdrop, a story about which employers contribute 10% versus 4.7% is not a story about retirement security. It is a story about the winners inside a system that was redesigned, between 1978 and the present, to produce exactly the inequality it now produces. The employers profiled are not doing anything wrong. They are competing for talent in a market where the defined-benefit alternative no longer exists for most new hires. The article is the problem — not because the facts are inaccurate, but because it presents the far-right tail of the distribution as the distribution, and treats the absence of the bottom half as an editorial choice rather than a structural fact.
The Secure 2.0 Act’s student-loan matching provision is, in this light, an acknowledgment that the system is broken — an attempt to patch the hole where debt service was cannibalizing retirement savings — dressed up as innovation. Employers are matching debt payments because wages are insufficient to cover both debt and retirement, and the policy response is a tax-code workaround rather than wages high enough to render the workaround unnecessary. This is the architecture the article celebrates without naming.
A retirement system that produces millionaires at Costco and nothing at 50 million households is not a system with a few good employers. It is a system designed to produce that outcome. The design is in the tax code — Sections 401(k), 408(k)(2) for SEP-IRAs, 403(b) for nonprofits, 457(b) for government workers — each structuring a tax deduction that scales with income, delivered through an employer intermediary that the worker does not choose. The worker whose employer offers nothing gets nothing. The worker whose employer offers 4.7% gets 4.7%. The worker whose employer offers 9% and who can afford to contribute the maximum and who never faces a hardship withdrawal becomes the subject of a feature in the national business press. This is what the numbers say. The methodology did not change. The score is the score.