The Inventor of the 401(k) Says It’s Broken.
Your Kids Have a Better Path.

In August 2026, Ted Benna — the benefits consultant who designed the original 401(k) plan in 1981 — told Moneywise that the system bearing his legacy has failed.[1] His critique is specific: the 401(k) disproportionately benefits higher-income workers, leaves lower-income earners behind, and needs to be replaced with something simpler. To try and fix these problems, he has proposed “Radish.”[2]

Adoption is still thin. As of late July 2026, large insurers, 401(k) recordkeepers, and a major university had all heard the pitch without one of them signing on, and pilots were running at a North Carolina private school, an Oklahoma retail operation, and a trucking firm of roughly 200 workers. In early August, Bloomberg reported that Radish had a couple of clients. No large employer has adopted it.

Benna’s frustration is worth taking seriously. However, for families with young children, the more important question is not whether he’s right about the 401(k). It is whether your children have to depend on an employer to build serious retirement wealth in the first place. They do not. And that is the entire argument this site was built on.

What Benna Actually Said

Benna’s complaint is structural, not political. The 401(k) is an opt-in system: a worker must have an employer who offers the plan, choose to participate, contribute from their paycheck, and stay long enough to be vested in any employer match. Each of those steps acts as a filter, and lower-income workers are more likely to fail at multiple points simultaneously.

The participation data bears it out, and the gap is wider than the headline numbers suggest. As of March 2025, 72% of private industry workers had access to an employer retirement plan and 53% participated. Among the lowest-paid tenth of workers, access was 38% and participation was 15%. Among the highest-paid tenth, access was 93% and participation was 83%.[3]

The number that actually explains it: who the employer is

Split that same survey by the size of the business and the national average dissolves. At private establishments with 1–49 workers, 55% had access and 38% participated. At establishments with 500 or more workers, 90% had access and 76% participated.[3]

Those participation figures are shares of every worker at businesses of that size, not of the ones who were actually offered a plan. Narrow it to the workers who had access and the enrollment rates are far closer together: 70% at the small establishments, 84% at the large ones.[3] That tells you where the 38-point gap really comes from. If small establishments offered plans at the large-establishment rate — 90% instead of 55% — and their workers enrolled at exactly the rate they already do, participation would rise from 38% to 63%. That is 25 of the 38 points closed by availability alone. Only the remaining 13 turn on whether a worker who was offered a plan decided to use it.

The Center for Retirement Research states the structural problem plainly: firms with fewer than 100 employees employ about a third of all private-sector workers, and only 49% of the smallest firms sponsor a retirement plan at all, against 98% of the largest.[4] This is the detail that gets lost every time the 401(k) is debated as though it were a universal system.

The reason is mostly arithmetic nobody has corrected. More than half of small employers believe a plan would cost over $10,000 a year to run, and nearly 30% put it above $20,000 — while the actual cost for a 25-person firm is frequently under $3,000.[4]

Benna’s alternative — Radish — is a 401(a) plan for workers earning under about $160,000 that the employer funds directly against performance targets: a few dollars for every shift started on time, a lump sum for hitting a safety or tenure goal. Because the money never runs through payroll, the employer saves payroll taxes on it, and the worker never has to opt in. It is a sensible idea, and it does remove the payroll-deduction step that causes so many lower-income workers to miss out. The problem is that it still requires an employer to act, and the employers who would have to act hardest are the ones already declining to sponsor a plan that costs less than $3,000 a year. A fix routed through the employer cannot reach a worker whose employer is not willing to make the investment in their workforce.

The Five Places the 401(k) Loses Lower-Income Workers

The failure mode is easier to see when you lay it out as a sequence. Radish addresses one of these five gates. The other four stay open.

• 1. The access gate. You cannot participate in a plan your employer does not offer. Workers in food service, retail, personal services, and seasonal trades — sectors that skew toward lower wages — are among the least likely to have access at all.
• 2. The employer-size gate. This is where the access gate actually bites, and it is the one most often left out. Low-wage work is concentrated in exactly the establishments least likely to sponsor a plan. The coverage gap is not sprinkled evenly across the economy — it sits directly on top of the part of the workforce that needs it most. A policy aimed at 401(k) participants is, by construction, aimed past them.
• 3. The participation gate. Opt-in systems produce opt-out behavior at the margin. A worker earning $35,000 a year who is told to defer 6% of their paycheck to capture the full employer match faces a real trade-off between a future benefit and a present expense. Many choose the present expense, and the match goes unclaimed.
• 4. The proportionality gate. Even a worker who clears the first three gates hits a fourth one nobody talks about. A match expressed as a percentage of salary pays the smallest dollar amounts to the youngest, lowest-paid employees — the exact people whose money has the most time to compound. Five percent of a 22-year-old’s $35,000 is $1,750. Five percent of a twenty-year veteran’s $120,000 is $6,000. The design routes the largest employer dollars to the people with the fewest years left to grow them. Nine out of ten of the highest-paid workers who are offered a plan put money into it. Fewer than half of the lowest-paid workers who are offered one do — 89% against 40%.[3] A match you cannot afford to claim is not compensation. It is compensation on paper only.
• 5. The leakage gate. Workers who do enroll often withdraw early when a financial shock arrives. In 2025, 6% of Vanguard participants took a hardship withdrawal — the highest share on record, up from 5% the year before.[5] A $5,000 hardship withdrawal can cost $1,500 or more in immediate taxes and the 10% penalty, and it forfeits whatever that money would have compounded into over the following twenty years.

The Government Match Arrives in 2027 — and It Skips Your Kids

Beginning with the 2027 tax year, the federal Saver’s Match replaces the old Saver’s Credit. Instead of a credit against a tax bill, the government contributes up to 50% of what an eligible saver puts in — a maximum of $1,000 per person per year — paid into their retirement account.[6] Two details get reported badly. The 50% is a ceiling, not a flat rate: it steps down as income rises through the phase-out. And nothing arrives during 2027 — the match is based on what you contribute that year, claimed on the 2027 return you file in 2028, and deposited after that return is processed. The full match runs up to $20,500 of modified adjusted gross income for single filers, $30,750 for head of household, and $41,000 for married filing jointly, phasing out entirely at $35,500, $53,250 and $71,000 respectively. MAGI is not salary — it adds back pre-tax retirement contributions, so deferring into a 401(k) does not pull you down into the range.

Two things about it matter for families. First, it works with a traditional or Roth IRA, not only a workplace plan — which is this post’s whole argument showing up in federal law. A worker with no 401(k) at all can still collect a federal match — provided the account can receive it. Neither plans nor IRA providers are obliged to accept Saver’s Match deposits, so the institution has to have opted in; that is a question worth asking before choosing where the account lives.

Second, and this is the part to plan around: an eligible person must be 18 by the end of the tax year, not claimed as a dependent on someone else’s return, and not a student.[6] “Student” here is the section 152(f)(2) definition — enrolled full time at a school during some part of each of five calendar months in the year — so a full course load through the spring semester alone is enough to rule the year out. Note what is not on that list: living at home does not disqualify anyone. Being claimed as a dependent does, and the two usually travel together, but they are different tests. So the eleven-year-old mowing lawns does not qualify, on age. Neither does the college sophomore working summers, on the student test. The match becomes available the year your child is out of school, off your tax return, and earning at the bottom of the scale — which is precisely the year almost nobody has heard of it. Put it on the calendar now, because a $1,000 federal deposit at age 23 is worth considerably more than the same $1,000 at 45.

The Accounts That Don’t Need an Employer

Here is the part that belongs in every family conversation that follows the Benna story: none of the accounts ELI covers require an employer to be involved at all.

The 401(k) sits at the far end of this timeline — after a child has grown, entered a workforce, found an employer who offers the plan, and decided to participate. Every other account in the table is available long before that moment. A child born in 2026 can have a Trump Account and a 529 by the time they are one month old. A child who earns $200 mowing lawns at age eleven can have a custodial Roth IRA before middle school ends. By the time that child’s first job comes with a 401(k) enrollment email, they will have had more than a decade of compounding already running.

The custodial Roth IRA is the one most families rule out too early, so it is worth being precise here. The IRS requires earned income to fund it — it does not require a W-2. Babysitting, lawn mowing, pet sitting, tutoring, and gig work all count. Keep a simple written log of date, amount, and who paid, and the contribution room equals what the child earned, up to the annual maximum. A child who earns $600 in informal summer work has $600 in Roth IRA contribution room, and the window to make a 2026 contribution stays open until April 15, 2027.

The 401(k) is an employer-gated system. The accounts above are parent-gated. That is the entire difference.

If you want to see which plan your child’s eventual employer is likely to offer — 401(k), SIMPLE IRA, SEP, 403(b), or nothing at all — the retirement plan types by employer reference lays out the whole landscape by employer type and size. It is a useful thing to read once as a parent, because it makes the small-employer gap concrete rather than abstract.

The Benna story also makes explicit a connection ELI has been drawing from the start. Wells Fargo’s 2026 Money Study found that 64% of U.S. parents with adult Gen Z children are still financially supporting them — rent, groceries, phone bills, insurance.[7] Benna’s critique of the 401(k) is a critique of a system that arrives too late, to workers who were never set up to use it well. The fix is not a better 401(k). It is starting before the 401(k) was ever relevant.

The Bottom Line

Ted Benna built the 401(k) to give American workers a retirement savings vehicle tied to their employer. Forty years later he is acknowledging what the data has been showing for years: the system works well for workers who have access, participate, get a match proportioned to a salary they have already grown into, and never face an emergency that forces an early withdrawal. That describes the top of the income distribution far better than the bottom — and it describes a large employer far better than a fifteen-person shop.

None of that is the problem ELI is here to solve. The problem ELI is here to solve is the gap that exists before the 401(k) ever enters the picture — the years between birth and a first paycheck when the compounding clock is running and most families have not turned it on yet. The accounts that close that gap are already available. They do not require an employer’s permission, a benefits enrollment window, or a company that has decided a plan is worth $3,000 a year. They require a parent who opens one.

References & Disclosures

1. Benning, Sarah. “Ted Benna built the 401(k) 40 years ago — now he says it fails lower-income workers and is pushing a simpler alternative.” Moneywise / Yahoo Finance. August 2, 2026. Yahoo Finance →
2. Radish Plan. Radish — an employer-funded 401(a) performance-incentive savings plan for workers earning under roughly $160,000, created by Ted Benna and co-founder Kyle Bagley from Benna’s 2024 “Wheat Grain Incentive Plan” concept. Product site, accessed August 21, 2026. IRALOGIX was named the plan’s IRA services provider on April 2, 2026. Adoption status from Michael Fischer, “The Father of the 401(k) Has a New Savings Plan,” WealthManagement.com, July 28, 2026 (pilots underway; no employer signed at that date), and Money, August 4, 2026, citing Bloomberg reporting that Radish had a couple of clients. radishplan.com →
3. U.S. Bureau of Labor Statistics. Employee Benefits in the United States, March 2025. Bulletin 2830, released September 25, 2025. National Compensation Survey; retirement benefit access, participation, and take-up rates for private industry workers, by establishment size and wage percentile. Reference period March 2025. BLS news release (PDF) →
4. Center for Retirement Research at Boston College. Small Business Retirement Plans: A Primer. August 11, 2026. Firm-size distribution of private-sector employment and plan sponsorship rates; small-employer cost perceptions drawn from the 2023 Small Employer Retirement Survey. Center for Retirement Research →
5. Vanguard. How America Saves 2026 (25th edition). Published June 16, 2026. Annual analysis of participant behavior across Vanguard-administered defined contribution plans; 2025 plan-year data. Note that Vanguard’s recordkeeping book skews toward large employers and does not represent workers without a plan. Vanguard →
6. Internal Revenue Service. Saver’s Match. Last reviewed August 14, 2026. Statutory eligibility, match rate, $1,000 annual cap and modified AGI phase-out ranges effective for tax years beginning after December 31, 2026; implementation guidance in Notice 2026-48. IRS →
7. Wells Fargo. 2026 Wells Fargo Money Study. March 30, 2026. Versta Research; 3,773 U.S. adults and 215 teens aged 14–17, fielded November 19–December 17, 2025. Wells Fargo Newsroom →

Early Life Investments is not affiliated with, endorsed by, or sponsored by Ted Benna, Radish Plan, Vanguard, Wells Fargo, the Bureau of Labor Statistics, the Center for Retirement Research, the Internal Revenue Service, or any other entity mentioned in this post. Statistics cited reflect published research as of the indicated report dates and are subject to change. This post is for general educational purposes only and does not constitute financial advice.