Half of American private-sector workers don’t have a workplace retirement plan at all — the 401(k), the most celebrated retirement vehicle in American life, simply does not exist for them as a workplace benefit. Among those who do have one, the average employer contribution is 4.7% of salary, and only 6% of plans offer a total match of 7% or above. These are the numbers the country has decided are normal. They are also, when you stare at them long enough, not a market outcome. They are a confession. The system produces 4.7% because that is what employers offer when nobody is making them offer more, and “nobody is making them” describes the working lives of most American adults.
Concession, since it has to be earned: the individual-responsibility people are not lying. Employees who do contribute, and who start early, and who stay with one employer long enough to vest, and who pick decent funds, materially benefit. A worker at Costco who puts in a decade plus and maxes out the match ends up with real money. The math works for the people for whom the math works. It is not nothing that the architecture rewards consistency. Fine.
Here is what the architecture also does, and what the brochures leave out.
Then there’s Stewart’s Shops, the regional gas-and-ice-cream chain, where employees average a 17% employer contribution and over 200 cashiers have become millionaires through stock ownership. There’s Ford and General Motors, where union workers negotiated their retirement contribution up from 6.4% to 10% in the 2023 contract. There’s Costco, where front-line hourly workers have built seven-figure accounts on a 4% non-elective contribution that rises to 9% after twenty-five years. Same country. Same tax code. Same 401(k) architecture. Wildly different outcomes. The variable that separates the 4.7% default from the 17% ceiling is not financial literacy, not individual discipline, and not the generosity of management. It is whether workers have power — through a union, through ownership, through a founder’s unusual conviction that keeping people is cheaper than replacing them, or through the scarcity of what they know how to do.
That last one is narrow by design, and it is worth naming. Boeing matches 10%. Visa puts in $2 for every $1 an employee deposits, up to 5% of pay — a 10% total on a 5% employee contribution. Mastercard does $1.67 per dollar on the first 6%, also landing at 10%. These companies aren’t responding to worker power in any collective sense. They are paying a premium for specialised talent in fields where the next employer will too. The leverage is individual, not structural — available to engineers, analysts, and product designers, unavailable to the people who stock shelves, drive trucks, or ring up groceries. The mechanism differs from a union contract or an ESOP, but the underlying logic is the same: the money follows whoever has the ability to walk away.
Stewart’s runs an employee stock ownership plan — an ESOP — instead of a conventional 401(k). Workers don’t just contribute to their retirement. They own a piece of the company that employs them. When it does well, their accounts reflect it. The 17% average contribution isn’t corporate generosity. It is the arithmetic of shared ownership, where profits that would otherwise flow entirely to outside shareholders flow partly to the people who stock the shelves. ESOPs are most common in manufacturing, construction, and engineering — sectors where a shrinking share of the American workforce is employed — and they remain a tax-favoured structure the federal government has chosen, by deliberate policy, to encourage. That choice can be widened.
Ford and General Motors arrived at a similar outcome from the opposite direction. Both automakers cut their defined-benefit pensions about two decades ago, replacing the old guarantee with a 401(k). Unionized workers, backed by the UAW’s bargaining leverage, negotiated the replacement into something worth having: a 10% nonelective contribution, up from 6.4%, won at the 2023 bargaining table. The number didn’t appear because executives had a change of heart. It appeared because a thousand workers had a counteroffer. Private-sector union density has collapsed to a small fraction of the workforce. The Ford-GM model is real, durable, and available only to the workers who still have a union to bargain with — which is why rebuilding sectoral bargaining coverage, the kind that lifts standards across an industry at once rather than one shop at a time, is the single most consequential labour-policy move on the table.
Costco sits between the retention model and the ownership model. Its founders built a philosophy around the idea that experienced workers cost less than constant turnover. The 4% non-elective contribution, rising to 9% for long-tenured employees, is part of that philosophy. It works. But Costco is neither worker-owned nor unionized. The model depends on the conviction of the people running the company. The moment a board decides quarterly returns matter more than turnover costs, the 9% is a line on a spreadsheet, and lines get cut.
Congress tried to help from the policy side. The Secure 2.0 Act, passed in 2022, lets employers match 401(k) contributions toward student-loan payments — Boeing, Verizon, Chipotle, and Comcast have already adopted the provision. It is a real expansion of the tools available, but it still depends on the employer choosing to be generous in the first place. The policy widens the door without pushing anyone through it. States are meanwhile auto-enrolling workers whose employers offer nothing at all — that is real and overdue. But the gap between nothing and 4.7% is the easy part.
The gap between 4.7% and 17% is where real retirement security lives, and closing it requires the thing that produces every good outcome in the economy when workers have it: organised, durable power at the point where the money gets divided. Three paths are sitting there waiting to be walked, and each has American precedent.
Multi-employer Taft-Hartley plans pool retirement contributions across an industry, so workers at small shops get the same bargained contribution as workers at the big contractors — the construction trades have run versions of this for decades. Expanded ESOP tax incentives, including the kind of broad-based S-ESOP credits Senator Sanders and others have proposed, would push ownership into the service sectors where it has barely reached. And sectoral retirement funds, the German model where an entire industry’s employers and unions jointly fund a portable pension that follows the worker from job to job, would do for retirement what sectoral bargaining did for wages. None of this requires inventing a new country. It requires deciding that the people who build a firm ought to share in what it earns — and writing the rules accordingly.
The country already has the proof. Stewart’s cashiers know it. The UAW members at Ford know it. Costco’s long-tenured workers know it, though they may not think of themselves as examples of a principle. The principle is simple: retirement security follows power. It always has. The question is whether the rest of the workforce is going to get some.